International Workshop Food Security in Complex Emergencies:

building policy frameworks to address longer-term programming challenges 23-25 September 2003, Tivoli, Italy

Workshop report

Volume 1: Main Report

Hosted by the Food and Agriculture Organization of the United Nations with funding from the European Commission under the joint EC/FAO project on Food Security and Complex Emergencies

Summary and synthesis

This report documents the proceedings of an international workshop on Food Security in Complex Emergencies: building policy frameworks to address longer-term programming challenges, held in Tivoli, Italy, 23-25 September 2003. The workshop was organized by FAOs Agricultural and Development Economics Division with funding under the joint EC/FAO project on Food Security and Complex Emergencies, and attracted 36 participants from 20 different intergovernmental, non-governmental and research organizations. The key workshop objective was to identify elements for policy frameworks to enhance the resilience of food systems in different complex emergency contexts and the role of information in supporting this. Discussion was structured around three themes: 1. Understanding the challenges complex emergencies pose for food security 2. Principles, approaches and policies for food security in complex emergencies 3. Implications for research and information systems, and next steps. Overall workshop conclusions Complex emergencies, involving governance failures and violent conflict exacerbated by HIV/AIDS and natural hazards, have profound impacts on food security and are rising in number and seriousness. Increasing flows of emergency humanitarian assistance have been dwarfed by falls in development aid, and needs for extended forms of assistance in protracted crises and for preventing and preparing for crises often remain unmet. This may partly be a result of the absence of policy frameworks for addressing longer-term food security issues in protracted crises. Neither the relatively well developed policy frameworks of humanitarian agencies, comprising principles such as humanity, neutrality, impartiality and independence, nor those of development actors which emphasize such principles as sustainability and cost recovery appear adequate for protracted complex political emergencies when people need support to carve out livelihoods for themselves over years or decades in highly volatile situations. This has been characterized as a policy gap: a lack of mechanisms and reference points for setting priorities in addressing food insecurity in the short and longer-term, based on people's needs and expectations. Embedding shorter-term measures in longer-term policy frameworks would allow their longer-term effects to be better taken into account. Requirements for strengthening such policy frameworks were identified, including development of organizational capacities for policy making in this area, establishment of appropriate ethical and operational principles for longer term programming in conflict, improved inter- and intra-sectoral coordination and beneficiary participation, and more generous, flexible and longer term funding arrangements. The likely efficacy of a wide range of policy options under different conflict conditions was examined. This highlighted the importance of strong knowledge-action links involving research and information systems which engage more effectively with complex political dynamics and social impacts of and responses to violence. More support to government information systems in crisis-prone countries is needed. Initial experiences with conflict early warning systems might offer useful lessons. Recommendations were made for future collaboration in policy development and research and information activities, aided by FAOs new information portal on food security in complex emergencies which was unveiled at the workshop. Summary by theme 1. Understanding the challenges complex emergencies pose for food security The number and seriousness of emergencies and protracted crises that result wholly or mainly from human factors such as conflict, macroeconomic collapse and HIV/AIDS is increasing. Violent conflict poses particular challenges arising from its capacity not only to disrupt food security and livelihoods but also to destroy or pervert the institutions and

authority structures on which recovery depends and thus to transform the socio-political and institutional context for aid. In case studies ranging from southern Africa to north-eastern DRC, the Karamoja region, Somalia, Eritrea, Ethiopia, Sudan and post-Taliban Afghanistan, participants showed how food availability, access and stability had been impacted in such situations. Availability losses from conflicts outweigh by far the provision of food aid. People have lost access to food through being directly affected by fighting, displaced, cut off from markets and relief supplies, deprived of employment and income or subjected to war taxes. Stability has suffered through impacts on prices and markets, loss of institutional support including for safety nets, and increased exposure to natural hazards. Donors have been obliged to respond by funding increasing volumes of humanitarian assistance, but this has been poorly allocated in relation to need and dwarfed by declines in development aid. Needs for extended forms of assistance in protracted crises and for preventing and preparing for crises often remain unmet. Participants identified four major challenges: a) Understanding complex emergency contexts as a basis for policy and programming. The highly differentiated and dynamic nature of complex emergency settings means there is little scope for generalized policy responses. Thus policy needs to be guided by a lightfooted and flexible research capacity, strongly linked to decision-making networks. Requirements and a framework for developing such a capacity were suggested. Political analysis is a prime requirement, encompassing the role of the state and of external actors, governance failures, spatial differentiation within countries, food insecurity as a tool of subjugation, and political influences on aid policy. Such political dimensions tend to be under-represented in information systems which feed into policy. b) Understanding the nature of vulnerability in complex emergencies. The concept of vulnerability is widely used to denote the susceptibility of people, systems or institutions to the impact of hazards, and has obvious relevance in analysing food security impacts in complex emergencies and designing action to mitigate those impacts. Yet the concept is subject to different definitions, indicators and methodologies in the realms of food security, development and conflict management. A consolidated vulnerability concept needs to be established for food security work in complex emergencies. A livelihoods approach to conflict vulnerability is helpful, but the sustainable livelihoods model needs modification to emphasize resilience, the capacity of a system to adjust and reorganize without losing its functionality, with vulnerability as endogenous to each component of the model rather than as an exogenous contextual variable. c) Interactions between complex emergencies, food security and HIV/AIDS. HIV/AIDS affects food security through impacts on agricultural labour supply and knowledge, household impoverishment and the sustainability of food production, macroeconomic effects and other risk factors which undermine the resilience of food systems. It is also linked to complex emergencies, for example through increased exposure among displaced populations, the behaviour of armies and militias, and pressure on women to resort to sex work when other livelihood options fail. Time-lags between HIV prevalence, AIDS deaths and inter-generational consequences such as increasing numbers of double orphans mean that impacts will continue to grow for at least another generation. d) Coping with the operational constraints imposed by political violence. There is a consistent tendency to underestimate the obstacles that political violence throws up for moving from relief to longer-term food security interventions. Programming is challenged by uncertainty and frequent reversals, inadequate assessment of needs and of the political situation, access to affected populations blocked by gatekeepers of various kinds, and problems of targeting, misappropriation, staff security and resources.

ii

2. Principles, approaches and policies for food security in complex emergencies Participants considered policy processes involved in the progression from contextual analysis to appropriate responses. A policy gap was identified in the sense that there is a lack of mechanisms and reference points for setting priorities in complex and protracted crisis situations to address peoples own critical food security concerns for the short and longer term. As short-term measures are very likely to have longer-term effects on food security, they need to be embedded in longer-term frameworks. Thus policy frameworks must be built which include appropriate principles, goals, approaches and yardsticks for this purpose. In practice, it was noted, agencies funding or implementing interventions in complex emergencies already have their own policy frameworks which condition their perception of those contexts and guide their policy-making. Yet as a basis for moving from short-term humanitarian assistance to food security oriented interventions with a longer-term horizon, these tend to be weak in several key respects: a) Interventions are often designed without passing through a deliberative policy process which examines current knowledge on the context to be addressed and evaluates options for programme objectives, methods, content and duration. As a result programme responses to food insecurity tend to be short term, standardized, inflexible and unresponsive to changing contexts. Programme design choices reflect specialized agency mandates and donor supply preferences, especially for food aid, while needs assessment exercises often take these choices as given. b) There is uncertainty over when and how it is practically and ethically appropriate to locate emergency humanitarian assistance within longer-term frameworks for helping to rebuild livelihoods, institutions and food security and to promote peace. International agencies increasingly strive to pursue all of these ends in protracted conflict situations such as in Somalia and Sudan. Yet compared with humanitarian principles, principles of development-oriented and resilience-focused work in such contexts are relatively underdeveloped. Furthermore, humanitarians have warned of the dangers of undermining the neutrality and impartiality on which access to affected populations depends, and that humanitarian assistance cannot be at the same time unconditional, as its principles demand, and contingent upon its ability to promote development and peace. c) Funding mechanisms and organizational and staffing structures for humanitarian responses are normally separate from those for development assistance, and there is uncertainty about where the boundary between the two lies. Activities which do not fall squarely in either category, aimed at supporting livelihoods and boosting resilience in food insecure communities, rebuilding institutions and services and mitigating conflict in protracted crises are poorly resourced. Several requirements were identified for strengthening policy frameworks:

Organizational cultures, structures and procedures need to change to strengthen policymaking capacities for food security in complex emergencies. While steps have been taken in this direction, for example in the recent establishment of FAOs Rehabilitation and Humanitarian Policies Unit for agricultural rehabilitation in emergencies, capacities need to cover all aspects of food security, link more closely with reconstruction and development structures and address prevention and preparedness issues.

Rather than attempting to establish in advance blueprint policies for generic types of
complex emergency situation, policy frameworks need to support a capacity for flexible, opportunistic and reiterative policy formulation matched to specific contexts and to the priorities of food-insecure people. Participants identified many examples of interventions, relating to agricultural and non-agricultural livelihoods, access to land and markets, nutrition, asset protection and safety nets, which might be effective under different conditions and which take conflict issues into account.

iii

 Stronger and more transparent ethical and operational principles are needed to guide
policy decisions on what forms of assistance beyond basic relief are appropriate under what circumstances, how and by whom they should be provided and to whom they should be accountable. The Red Cross Codes of Conduct and new Sphere standards on food security may provide a useful starting point for these. Participants agreed on the need to seek opportunities for short and longer-term interventions to support livelihoods and food security. Policies must take into account not only how conflict constrains implementation, but also how interventions are likely to be incorporated into conflict and the ethical and practical implications of this.

Bringing about peace was proposed as an important element to be considered in food

security policy frameworks, as the aim of protecting food security in complex and protracted crisis situations is related to the political aim of ending conflict. The argument for such a primacy of peace imperative is based on the fact that conflicts have direct victims, that direct destruction and the prolonged effect of distorted social and economic processes destroy assets and have other long-term effects, and that its indirect impact can generate famine and food insecurity, making recovery extremely difficult. Giving primacy to peace building concerns requires adequate knowledge of the political economy, clarity about aims and about how interveners relate to humanitarian actors, a central role for the UN in legitimising interventions, and recognition that complex crisis can bring opportunities for positive change.

While policy frameworks are likely to be agency-specific, elements of them can be shared.
Inter- and intra-sectoral coordination among agencies requires further strengthening to ensure sharing of knowledge, coherence in dealings with authorities and factions, flexibility and capacity to plan for different scenarios, and effective linking between short and longer term intervention strategies. Local and regional inter-agency networks, both formal and informal, and upgrading of local staff were highlighted, as was the importance of transparency and local participation in policy processes.

The effectiveness of policy frameworks will depend on donor commitment to more

generous, flexible and longer-term funding arrangements for food security interventions in complex emergencies. 3. Implications for research and information systems, and next steps Participants stressed that policy frameworks for food security in complex emergencies need to support strong knowledge-action links involving research and information systems that adequately cover social and political dimensions of changing contexts as well as the standard food security indicators. In this regard it was noted that: a) Security conditions in complex emergencies preclude monitoring of many indicators, and this increases the importance of background knowledge and opportunistic information strategies such as surveys among displaced populations. b) Government-run food security and nutrition information systems in many crisis-prone countries are weak and gain little benefit from externally provided, state-of-the-art but temporary information support for emergency relief programmes. Some of the resources available in emergencies could profitably be used to reduce vulnerability to future crises by boosting the ability of government information systems in these countries to promote a better baseline understanding of livelihoods and how they are changing. c) Information systems in complex emergencies are effective in providing data and information to meet the technical needs of most humanitarian operations, but generally fail to engage with the deeply complex problems that concern affected populations. Longerterm food security programming requires a greater focus on knowledge, e.g. on how societies are affected by and demonstrate resilience in the face of violence. d) Conflict early warning systems are presently at a rudimentary stage, and face unresolved issues of political sensitivity and appropriate methodologies and indicators. Their further

iv

development needs to be supported, and consideration given to their potential role in linking with research to identify opportunities to mitigate or prevent conflict as well as assessing threats, and their links with food security information systems. Follow-up action recommended at the workshop included ways in which intervening agencies can further define policy frameworks for longer term food security in complex emergencies. Participants endorsed and provided feedback on FAOs new international portal on food security (http://www.fao.org/crisisandhunger), which will provide access to information and online networks to field practitioners and food security and policy analysts involved in designing broader and longer-term policy frameworks. An agenda for future research was identified for improving understanding of links between food security and conflict and of the institutional and policy architecture for food security in complex emergencies. Participants expressed a keen interest in future collaboration with FAO, both individually and institutionally, to build on this workshop and make further progress in this field.

Contents
Summary and synthesis Contents Acknowledgements Glossary Introduction The rationale for the workshop Workshop objectives, themes and participation Structure of this report Welcome and keynote address i vi vi vii 1 1 2 2 2

Theme 1: Understanding complex emergencies and the challenges they pose for food security 7 Session 1: Complex emergencies and food security challenges 7 Session 2: Complex emergency processes, challenges and programming constraints experiences in different contexts 16 Theme 2: Principles, approaches and policies for food security in complex emergencies (Session 3) Introduction Plenary panel How risk of violent conflict affects policy choices for supporting food security Theme 3: Implications and next steps Introduction Session 4: food security information systems in and out of crisis Session 5: Conclusions on principles and agendas for action Annex A: Policy - conflict matrix 26 26 26 30 48 48 48 53

Acknowledgements
The FAO management group for the workshop, under the overall direction of Prabhu Pingali, Director of Agriculture and Development Economics Division (ESA), comprised the following staff of the Food Security and Agricultural Projects Analysis Service (ESAF): Margarita Flores (ESAF Chief), and Luca Alinovi (Project Coordinator), Gnter Hemrich (Food Security Analyst), Jacky Sutton (Information & Communications Consultant) and Sandra Stevens (Workshop Secretary) of the FAO/EC Project "Improving Early Warning and Food Security Information in Countries Subject to Complex Emergencies". Other FAO staff acting as daily rapporteurs were Richard China (Emergency Operations and Rehabilitation Division) and Mark Smulders and Christian Lovendal of ESAF. Philip White acted as Lead Consultant for the workshop and prepared this report. Ms Eunice Shankland, Facilitation Consultant, provided preparatory advice on facilitation methods and processes. Particular thanks are due to participants from outside FAO who gave up their time to prepare workshop papers and spend three days in intensive discussion in Tivoli. Two of the participants, Kate Longley and Franois Grunewald, kindly agreed to travel early to Rome to assist as workshop facilitators as well as chair sessions. Other sessions were chaired by Stephen Devereux and Clement Obonyo Jabwor.

vi

Glossary
ALNAP CBO CEWARN CIA CSI CSO DAC DFID DPRK DRC EC EPLF ESA ESAF FAO FEWS NET FSAU HIPC ICRC IDPs IFPRI IGAD IHL INGO IPALMO LRA NGO NMPACT OCHA ODA ODI OECD OFDA ORH PIPs PRSP RENEWAL SACB SC-UK SLF SPLM UNDP UNOSOM UPC UPDF USAID VAM WFP Active Learning Network for Accountability and Performance Community-based Organization Conflict Early Warning System US Central Intelligence Agency Coping Strategies Index Civil Society Organization OECD Development Assistance Committee UK Department for International Development Democratic Peoples Republic of Korea Democratic Republic of the Congo European Commission Eritrean Peoples Liberation Front FAO Agriculture and Development Economics Division FAO Food Security and Agricultural Projects Analysis Service Food and Agriculture Organization of the United Nations Famine Early Warning System Network (USAID) Food Security Assessment Unit Highly Indebted Poor Country International Committee of Red Cross & Red Crescent Societies Internally Displaced Persons International Food Policy Research Institute Intergovernmental Authority on Development International Humanitarian Law International non-governmental organization Institute for Relations between Italy and Africa, Latin America and the Middle East Lords Resistance Army Non-governmental organization Nuba Mountains Programme Advancing Conflict Transformation UN Office for the Coordination of Humanitarian Affairs Official Development Assistance Overseas Development Institute Organization for Economic Cooperation and Development Office for Foreign Disaster Assistance (USAID) Operation Restore Hope (Somalia) Processes, Institutions and Policies (in the SLF) Poverty Reduction Strategy Paper Regional Network on HIV/AIDS, Rural Livelihoods and Food Security Somalia Aid Coordination Body Save the Children (UK) Sustainable Livelihoods Framework Sudan Peoples Liberation Movement United Nations Development Programme United Nations Operation in Somalia Union of Congolese Patriots Uganda Patriotic Defence Force United States Agency for International Development Vulnerability Analysis and Mapping UN World Food Programme

vii

Introduction
This report documents the proceedings of a 3-day international workshop entitled Food Security in Complex Emergencies: building policy frameworks to address longer-term programming challenges, held in Tivoli, Italy, from 23rd to 25th September 2003. The workshop was organized by FAOs Agricultural and Development Economics Division (ESA) with funding under the joint EC/FAO project on Food Security and Complex Emergencies.

The rationale for the workshop

The rationale for the workshop was the need to respond to two main recent trends affecting food security:

The number and scale of conflict-related food security emergencies is increasing, and the
role of violent conflict in escalating a natural hazard, such as a drought, to a food security emergency has grown over the last decade. Linked to both natural and human-induced crises is the spread of HIV/AIDS.

While donors are increasingly obliged to fund short-term emergency interventions, needs
for extended forms of assistance in protracted crises and for preventing and preparing for crisis often remain unmet, and resources for long-term development aid have stagnated or decreased. More than half of the countries where undernourishment is most prevalent experienced conflict during the 1990s. Countries subject to complex emergencies are characterized by weakness or absence of government functions, poorly performing institutions, authorities which lack legitimacy, and a tendency for any transition from violent conflict to peace to be protracted, uncertain and prone to reversal. Profound and multiple impacts on livelihoods and food security often result from complex and highly context-specific interactions involving such factors as environmental hazards, resource scarcity and degradation, political and socioeconomic marginalization, lack of basic services, crime and the spread of HIV/AIDS as well as direct war impacts such as insecurity and displacement. International responses to complex emergencies concentrate largely on emergency relief, in particular food aid, and short-term agricultural assistance. Information systems and activities in support of these interventions focus on quantifying relief needs, beneficiary targeting and logistical aspects. Yet complex emergencies may, in their own distinctive ways, persist for many years in a state somewhere between all-out war and peace within which conditions may vary widely over space and time. Populations living in the midst of such chronic but fluctuating conflict strive to protect their livelihoods and food security as best they can, whether in receipt of emergency humanitarian assistance or not. Within the limits of a short-term planning horizon, emergency agricultural support programmes such as those implemented by FAO address these needs in increasingly diverse and imaginative ways, some with significant capacity-building effects. The question arises as to whether these short-term responses can be located within broader and longer-term food security policy frameworks aimed at improving peoples resilience, preparing to take advantage of lulls in conflict or the prospect of a more sustained peace, and helping prevent further disasters. Humanitarian and development agencies alike have begun to converge on the search for new policy frameworks of this kind, but their introduction faces a number of constraints:

It requires a concomitant broadening of capacity to monitor and analyse trends in food

security, livelihoods and the political economy and their internal and external determinants, as well as up-to-date assessment of humanitarian and security developments. Yet most agencies lack access to the breadth and depth of information as well as the institutional memory or expertise to support the necessary analytical work.

 Agency mandates and funding cycles, and inter-agency coordination mechanisms, rarely
permit the required degree of flexibility for them to respond to rapidly changing circumstances.

To the extent that it implies increased involvement of authorities (governments or rebel

organizations) which are party to ongoing, systemic political violence, it can undermine agency efforts to adhere to humanitarian principles of independence, neutrality and impartiality. For these reasons, a strategy often advocated in natural disasters of blurring the distinction between humanitarian and development assistance cannot be uncritically extended into the realm of complex emergencies. These factors, alongside the political conditionality often applied by donors, have caused not only uncertainty over what broader support to livelihoods and food security is appropriate under what circumstances, but also weaknesses and delays in immediate emergency response. Thus while humanitarianism is guided by a clear set of principles, concepts for longer-term policies and interventions require further development.

Workshop objectives, themes and participation

The workshop aimed to bring together participants who have worked on food security issues from a range of perspectives UN, government, NGO, civil society, donors and research in different parts of the world to:

explore food security challenges in complex emergencies; identify elements for policy frameworks to enhance the resilience of food systems in
complex emergency contexts;

find ways to strengthen food security analysis, information and research in such contexts; promote continued debate among a variety of actors.
Accordingly the three days of the workshop were devoted respectively to three main themes: 1. Understanding the challenges complex emergencies pose for food security 2. Principles, approaches and policies for food security in complex emergencies 3. Implications for research and information systems, and next steps. In the event, 36 people participated in the workshop, representing 20 agencies of all the various kinds mentioned above. Many were internationally recognized figures in their respective agencies and/or in research on food security, livelihoods and conflict.

Structure of this report

This report is in two volumes. Volume 1 provides an account of proceedings under each of the above themes based on plenary discussions and working group reports, with reference also to main points of papers prepared by participants. Annex A presents a summary matrix table relating policy issues to conflict scenarios resulting from the Theme 2 group exercise. Volume 2, under preparation, is a compilation of papers that participants were invited to prepare for presentation at the workshop.1

Welcome and keynote address

In welcoming participants and explaining objectives Margarita Flores, Chief of FAOs Food Security and Agricultural Projects Analysis Service (ESAF), noted that the Committee on World Food Security had raised the issue of the impact of conflict on increasing food insecurity. The Workshop, she said, would provide elements of discussion on food security policy for the CFS.
1

These include papers by two prospective participants who were unable to attend the workshop.

Keynote address: Food security and complex emergencies challenges for FAO FAOs major concerns in this area were set out in a keynote address by Prabhu Pingali, Director of FAOs Agriculture and Development Economics Division (ESA), providing graphic evidence of the trends referred to above. The increase in the number and proportion of emergencies attributed wholly or mainly to human factors, mainly violent conflict but also economic crises, is striking. Hunger and conflict tend to coincide around the world, and especially in Africa. The links between the two are multiple and complex. Yet the international response has been disappointing. While assistance for emergency humanitarian interventions has mushroomed since the early 1980s, this has been dwarfed by the dramatic fall in official development assistance since the end of the Cold War (Figure 1). These developments were examined in relation to the Democratic Republic of Congo (DRC), Somalia and Sudan, demonstrating further that humanitarian assistance has itself not been provided evenly or optimally relative to needs across countries, affected groups or time. Prabhu Pingalis presentation showed how conflict has impacted on each of the three main dimensions of food security that FAO distinguishes: availability, access and stability. Availability losses from conflict, concentrated in Africa and West Asia, outweigh by far the provision of food aid. People have lost access to food through being directly impacted by fighting, displaced, cut off from markets and relief supplies, deprived of employment and income or subjected to war taxes. Stability has suffered through impacts on prices and markets, loss of institutional support, including for safety nets and increased exposure to natural hazards. Many of these food security excl. emergency assistance emergency assistance impacts can also result from ODA billion US$ billion US$ Developing countries epidemics such as HIV/AIDS and 100 10 macroeconomic policy failures. 90 9 HIV/AIDS, for example, impacts 80 8 on labour, productive capital and End of Cold War 70 7 agricultural knowledge 60 6 50 5 (availability), on household 40 4 disposable income and 30 3 traditional safety nets (access) 20 2 and on viability of institutions and 10 1 markets (stability). Economic 0 0 collapse has profound impacts 73 75 77 79 81 83 85 87 89 91 93 95 97 99 01 on food import capacity and ODA excl. emerg. Assistance emergency assistance technology (availability), on Source: FAO-ESA employment and safety nets (access), and on markets and Figure 1: ODA and Emergency assistance in prices (stability). It is not developing countries surprising that many including some of the participants at this workshop would like to see the definition of complex emergencies broadened to include emergencies which do not involve war but do result from complex interactions between hazards of different kinds: epidemiological, governance, economic and environmental; longer term as well as short term. From the policy point of view these developments throw up four main challenges for FAO: 1. Gaining a better understanding of the factors that contribute to the resilience of agricultural and food systems in protracted complex emergencies; 2. Developing new approaches to designing flexible, principled support for that resilience in situations subject to political manipulation and rapid change; 3. Establishing responsive policy and planning frameworks capable of putting these approaches into effect; 4. Making sure that these frameworks use field research and information systems that can
3

adequately capture the complexity of complex emergencies. Overview of workshop rationale and process Philip White, Consultant to the Food Security and Agricultural Projects Analysis Service, outlined the rationale and proposed process of the workshop by reference to a simplified model of the policy formulation process for food security in complex emergencies shown in Figure 2, the essential logic of which is as follows:

Complex emergency contexts encompass a range of causal dimensions and short and
longer-term food security impacts as set out in the keynote address. People engage in and respond to these impacts (whether as perpetrators, beneficiaries or victims of violence) in ways which influence the nature and evolution of the crisis itself, as do neighbouring states, world powers and global networks pursuing their respective political and economic interests. These contexts offer both needs/opportunities for and constraints to external assistance aimed at helping people survive and cope better.

The process by which international actors formulate policy to support food security in
countries affected by complex emergencies is influenced by two sets of factors:
One is the range of principles and approaches which guides policy formulation

(represented by the shaded box in Figure 2) and which acts as a lens through which the complex emergency context is perceived. These include not only humanitarian, conflict transformation, livelihood-oriented and development principles and funding mechanisms, but importantly also foreign policy objectives of bilateral actors.
The other is the assessment of the context itself, made possible by an array of

research and information systems providing background statistics and knowledge on the country and affected areas, shorter term political and food security assessments, and up-to-date information on the humanitarian situation. These include international media networks which are an important driver of policy.

The outcome of policy formulation is a response programming exercise which results in a

collection of intervention projects of different duration and design. Apart from policies, programming is also shaped by:
a programming needs assessment exercise which makes use of the output of

information systems;
the results of evaluation of existing and past projects.

Intervention projects in turn influence the nature and dynamics of the complex emergency
context, hopefully in ways that reduce vulnerability and increase resilience for the most severely affected groups, but by virtue of introducing resources into a politically charged and resource-constrained environment inevitably also becoming part of the dynamics of the conflict itself.

The solid and dashed arrows in Figure 2 suggest that this process tends to have stronger
and weaker links:
Research and information systems are normally geared towards providing timely

assessments of the dimensions of crisis in terms of evolution of conflict and other hazards, of their food security impacts, and to some extent of ways in which people respond and try to cope. They are less adept at identifying opportunities for assistance and constraints.

Violent conflict, natural hazards, HIV/AIDS, economic collapse etc.

Food security impacts

Complex
Priorities and responses of affected groups Opportunities for assistance

Emergency

Constraints to assistance

Context

Principles & approaches guiding policy Humanitarian, conflict transformation, capacity building, reliefdevelopment linking, foreign policy, funding mechanisms etc. Evaluation Food security policy formulation for country or affected area Needs assessment exercise Stronger links: Projects < 1 year Projects 1-3 years Projects 3-5 years Weaker links:

Evaluation

Response programming

Figure 2: From context to food security responses in complex emergencies: a process model
The link between contextual assessments by research and information systems and

policy formulation also tends to be weak. There is a tendency for policy responses for food security to be standardized, inflexible and unresponsive to the changing nature of the emergency, often concentrating on external procurement of food aid, seeds and tools.
The programming of responses tends therefore to be guided more by a needs

assessment exercise which takes the nature of the response (a policy issue) as given

Research & information systems for food security

International & regional links

Dimensions of crisis

and focuses primarily on identifying what balance of standardized interventions is required for what groups, where, in what quantities and for how long.
Projects that result tend therefore to be vehicles for short-term, commodity focussed

interventions, usually of less than a year in duration and dominated by food aid and seeds and tools provision.
Evaluation processes focus on the efficiency and timeliness of these interventions and

feed into response programming. In general they only weakly address the appropriateness of policies responsible for designing the response. They also tend to neglect the usefulness of existing research and information systems in policy design. This generalized representation of the issues offers a conceptual framework for considering why aid interventions in protracted complex emergencies are dominated by short-term, highly standardized designs which are often insensitive to peoples longer-term food security needs. Recently, more livelihoods-oriented programming in such contexts has expanded in promising new directions, as participants contributions on Somalia and other cases illustrate. Yet these remain largely piecemeal and often beset by ethical dilemmas related to the conflict and the need to engage with belligerents, and their impact in terms of food security outcomes not fully evaluated. The principles and approaches guiding policy formulation elements of what is referred to as a policy framework in the workshop title are not all mutually compatible and are the subject of a number of ongoing debates, especially when it comes to conditions and requirements for moving from short-term emergency interventions to longer-term food security ones in situations of protracted instability.

Theme 1: Understanding complex emergencies and the challenges they pose for food security
Session 1: Complex emergencies and food security challenges
Introduction This opening plenary session aimed to build a common understanding of what we mean by complex emergencies and what kinds of challenges they pose for food security. Presentations by three panels, representing perspectives from the field, from international agencies and from research, were followed by plenary discussion. Panel 1: Perspectives from the field This first workshop panel offered three views of the local realities of working in complex emergency contexts. Somalia: the disputed regions of Sanaag and Sool Fatima Jibrell, Director of Horn Relief, an NGO working with pastoral communities in the area, opened with an account of the situation prevailing in the region of Sanaag in northeastern Somalia. Ninety-two percent of the population of Sanaag are pastoralists. With livelihoods structured around mobility they tend to have little respect for regional or international boundaries. Sanaag, along with its neighbouring region Sool, is facing a humanitarian disaster after three consecutive years of drought. These two regions lie between the semi-autonomous State of Puntland in the north-east and the self-declared State of Somaliland in the north-west, and are claimed by both. The areas relative peace since 1991 has attracted thousands people displaced by conflict elsewhere in Somalia, leading to a doubling of its predominantly pastoralist population. This, combined with the ban on livestock exports to the Gulf, has intensified pressure on rangeland resources to the point where, with increasing distances between water points and adequate pastures, pastoralist livelihoods have become unsustainable and some 70 000 people in eastern Sanaag are threatened with starvation. Pastoral communities have turned to cutting acacia for charcoal production as a source of income, exacerbating rangeland degradation. Water is increasingly scarce and 80 percent of camels have been lost. Nutrition status has declined, measles and diarrhoea have raised child mortality. Many families are moving into Bossaso. The international community had been unnecessarily slow to react to the mounting food security crisis in eastern Sanaag despite a series of early warning reports based on joint assessments by Horn Relief and the FAO-run Food Security Assessment Unit for Somalia. This is in part because of the complex political context. The crisis area is dominated by a clan affiliated with the Puntland administration, yet UN agencies, rather than approaching affected communities directly, have tended to work through political leaders who are part of the Somaliland administration which tends to view these pastoralist communities with suspicion. Neither administration had declared a drought emergency in the area, and international agencies had vacillated about the need for emergency response pending the organization of another interagency assessment of the food and security situation. There is a need not only for urgent humanitarian assistance but also for a concerted, longerterm international effort to address food insecurity in Sool and Sanaag based on grassroots consultation with pastoralist communities concerning the local political and security situation as well as livelihoods and food security.

The Karamoja cluster Sam Kona, attached to FEWS NET Nairobi, spoke of the Karamoja cluster of pastoralist communities living in the region bordering southern Sudan and Ethiopia and northern Uganda and Kenya, and subject to a range of conflicts. For over four decades, this area has experienced persistent and increasing insecurity, unrest and instability due to acquisition of modern small arms by mainly pastoral groups in the area. This has escalated cattle raids, armed conflict and road thuggery throughout the area. The unchecked gun trafficking from Sudan and Kenya has compounded and complicated the security situation and development efforts and kept the region in a state of impoverishment. The fact that conflict is a source of livelihood in the area has proved a problem for agencies seeking to provide appropriate relief and development interventions. Providers of humanitarian assistance tend not to address conflict and often end up exacerbating it. A few agencies are pursuing a conflict resolution agenda, but tend not to address the deeply entrenched roles of conflict within local livelihood systems. The question was posed: how can humanitarian, development and conflict resolution agencies work meaningfully together? Emergency agricultural operations: working at the coalface Cristina Amaral, Senior Operations Officer in FAOs Emergency Operations and Rehabilitation Division, provided some sharp reminders of the practical problems of accessing affected populations, obtaining adequate information and ensuring timely delivery in physically insecure areas where local institutions have collapsed. These problems mean that it is often simply not possible to undertake interventions at the most appropriate time or at the peak of a crisis. There is pressure to deliver, little time for planning and reflection, and necessary instruments and methodologies for rapid assessment of needs and of the political situation are often lacking. The aim of FAOs emergency rehabilitation work is to make food aid redundant by providing beneficiaries with the means to recover production. But there is a need to understand political realities, in particular the role of gatekeepers e.g. warlords, institutions, funding agencies and local authorities who can circumscribe access to affected populations. Where security is a major constraint and humanitarian actors have to be protected by armed guards, participation is wishful thinking, but there is a need to find mechanisms to dignify those who help. In protracted conflicts, programming is seriously challenged by uncertainty and frequent reversals which make it very difficult to think long term. In Afghanistan during 1992-94 longer-term programming got under way, but ten years later there are still acute humanitarian as well as rehabilitation and development needs, while the informal economy has become dominated by guns and opium. Donors need to start to engage in conflicts earlier on, at the stage when institutions are starting to decay and collapse. The main imperative is to work fast and stick to simple, practical solutions. Panel 2: The international agency perspective Analysing vulnerability in conflict Presenting a paper prepared jointly with Umberto Triulzi and Pierluigi Montalbano2, Massimo Tommasoli of the Italian Ministry of Foreign Affairs, Directorate General for Development Co-operation assessed the approaches taken by main international aid organizations in analysing vulnerability. Vulnerability has become a key concept in denoting the susceptibility of people, systems of institutions to the impact of hazards and in designing action to mitigate those impacts. Yet depending on both the different sectors from which it is derived and the
2

Umberto Triulzi and Pierluigi Montalbano both work at the Institute for Relations Between Italy and Africa, Latin America and the Middle East (IPALMO)

types of negative outcome at issue, vulnerability is subject to a wide range of definitions, indicators, measurement tools, methodologies and time frames. These are paralleled by a vast array of conflict assessment methods which attempt to gauge risk factors leading to contemporary conflicts and their adverse impacts on development. Broadly there are three constituencies in approaching vulnerability:

the food security family donors concerned primarily with development policy and socio-economic analysis conflict experts.
FAO, WFP and USAID, with NGO partners, have taken the lead in initiatives to progress vulnerability analysis in the area of food security, though more needs to be done to integrate analysis of exogenous risks, including macroeconomic ones, with that of socio-economic characteristics of affected groups. The World Bank and UNDP have developed a detailed methodology for vulnerability analysis focussing on the micro level, though widening this to the macro scale to include effects of trade policy and globalization processes on equity and well-being remains a challenge. The DAC has also played a key role in conceptualising vulnerability and recognizing its relevance to poverty reduction. Vulnerability to conflict is a newer area of study into which socio-economic vulnerability analysis needs to be properly integrated: recent approaches, for example by DFID, examining two-way links between conflict and development are a step in this direction. Tommasoli saw a need to consolidate these scattered and disconnected traces of vulnerability analysis by incorporating socio-economic vulnerability analysis into traditional assessments of conflict. This would not only provide a more accurate and preventionoriented analysis of fragile states, but could also improve coordination between development cooperation and conflict prevention. Socio-economic vulnerability analysis addresses the long-term objectives of development cooperation and, at the same time, promotes a forwardlooking and preventive outlook consistent with conflict prevention strategies. It also provides a common and shared methodological platform for development cooperation policy-makers and acts as a direct link to conflict prevention strategies. He suggested three areas for attention if the potential of socio-economic vulnerability analysis as a missing link between conflict prevention and development cooperation is to be realized:

develop a wider, less sectorally divided form of food security and socio-economic
vulnerability assessment which builds in aspects of vulnerability related to violent conflict and social and political superstructures and institutions;

conversely, explore ways to integrate the socio-economic dimension more squarely into
conflict prevention strategies;

develop consensus on a more comprehensive vulnerability assessment methodology in

conflict prevention strategies and development cooperation policy. Food aid and livelihoods: challenges and opportunities in complex emergencies Allan Jury, Chief, Food Security, Safety Nets and Relief Service in WFPs Division of Policy, Strategy and Programme Support, presented a paper prepared by Valerie Guarnieri, Senior Policy Analyst in the Division3. WFPs operation has expanded over recent years from US$1 billion to US$3 billion annually. From the classic conflict emergencies which predominated in the mid-1990s, attention is shifting to the increasing incidence of combined natural-conflict related emergencies such as in southern Africa and the Democratic Peoples Republic of Korea (DPRK). Recognition is growing that food aid and other assistance can save more lives in the long term if it is provided to those whose livelihoods are at risk as well as to those already destitute. But this broadening of the traditional focus of humanitarian
3

Valerie Guarnieri was on maternity leave at the time of the workshop.

assistance is not without its challenges and is not always appropriate. There is a consistent tendency to overestimate the extent to which it is possible to move from relief to recovery. In 2002, WFP assisted 44 million people through its emergency operations (EMOPs) and 14 million through its protracted relief and recovery operations (PRROs). Almost half of WFPs total emergency assistance was provided in complex emergencies, and included food distribution; supplementary feeding support; food for work; school feeding; food for training; and support for demobilization. Operational challenges include:

Humanitarian access: constrained by fighting, insecurity, denial or obstruction, logistical

problems

Directing the right aid to the right people: constrained by manipulation by warring parties,
reducing reliability of data, e.g. on numbers and needs; also by lack of complementary assistance

Preventing misappropriation: theft, post-distribution taxation, inflated registration or

exclusion of marginalized groups, leakage. Strategies include ground-rules and changing distribution modalities.

Ensuring staff security: over 230 UN staff members have died from malicious acts (direct
attacks, mines, threats, intimidation, kidnapping etc) since January 1992. Safety concerns may be used as an excuse to restrict humanitarian assistance operations in key areas.

Managing relief: absence of government or civil society institutions reduces accountability,

requiring substitute mechanisms, e.g. appointment of a Humanitarian Coordinator, designation of a lead agency, creation of a UN Transitional Administration

Limited resources: affect food/non-food mix; in protracted crisis donor fatigue can set in.
Allan Jury illustrated these challenges by reference to northern Uganda, where a post-war recovery strategy has been disrupted by the Lords Resistance Army (LRA). There has been no chance to test the effectiveness of this strategy. As Jury put it, ten dollars worth of LRA disruption can wipe out a hundred thousand dollars worth of investment. This is an example of how an assumption of a linear progression from relief to recovery in such contexts has become discredited. Protecting and supporting livelihoods as an emergency response has several benefits: safeguarding productive capacity, building recovery into the response, helping reduce relief dependency, cost-effectiveness, participation and so on. But there are limitations: a) Any humanitarian assistance especially food aid introduced into a resource-strained environment can play into the dynamics of conflict. These challenges are amplified with livelihoods support. A clear understanding of the political/military context is critical. Striking an appropriate balance between working with and working around power bases is difficult. b) Refugee and some IDP situations offer both constraints and opportunities for supporting livelihoods. People in camps are insecure, depend on alternative livelihood sources, and face problems of livelihood continuity on return home. But a UNHCR-led common strategy for food and non-food assistance (e.g. the WFPUNHCR Memorandum of Understanding) can permit intersectoral linkages essential for livelihood interventions. Most camps have functioning markets and some labour opportunities (within or outside) to support livelihoods. c) Food aid is not always appropriate for preserving assets or supporting livelihoods. It is important to gauge its impact on policies, institutions and processes that influence livelihood strategies, especially markets. Cash interventions may be preferable. The implications of these challenges are that WFP efforts to use food aid to support livelihoods need to:

10

 understand these risks by analysing violence and developing protection strategies; link pre-emergency interventions to emergency response to build resilience to hazards; integrate livelihood and emergency needs assessments, including within households; improve timing of interventions: incorporate livelihood risks into early warning efforts; advocate for those at risk of losing livelihoods whether food aid is appropriate or not; build strong partnerships and improve assessment, design and implementation skills.

Consensus among donors on analysis and indicators Jean-Claude Esmieu, Unit Coordinator (F5) of the EuropeAid Co-operation Office, spoke of the defencelessness of the international assistance community in the face of increasing conflict. Donors try to optimize the use of limited resources by choosing theatres of operations in which aid will be most effective, but there is a need to employ better indicators and analysis and build inter-agency consensus to improve the credibility of appeals. In the DPRK, for example, the analysis of FAO and WFP tends to differ in ways that reduce the confidence of the EC and other donors. Donors tend these days to insist on a system of indicators of impact along different parts of the crisis cycle. Yet the EC tends to have one analysis, its member states have others, and the US has yet another. This militates against optimization of use of funds and effectiveness of response. Donors also have to relate to recipient governments or authorities who have different agendas again, not always humanitarian. Donors cannot therefore be expected to bear all the responsibility for aid shortcomings governments must shoulder some of it. On managing conflict and building peace, donors tend to be even more at odds. There is a need therefore for a better consensus between donors on responding to complex emergencies, including on analysis. Panel 3: The research perspective Appropriate institutional responses: the centrality of an analytic capacity While the UN defines a complex emergency as a humanitarian crisis in which internal or external conflict causes a breakdown of authority and which requires an international response that goes beyond the mandate or capacity of any single agency and/or the ongoing UN country programme4, this says more about UN institutional responsibilities than about the situation per se. Martin Doornbos, Emeritus Professor at the Institute for Social Studies in The Hague, stressed that complex emergencies is a catch-all phrase for a highly differentiated set of contexts, each with its own history of causation. Some basic characteristics are that they are political, involving war or endemic political violence so that liaisons with government departments may be very problematic; they may involve grey time zones between chronic food deficit areas and protracted crises resulting from them; and they are likely to bring about new, qualitatively different situations. Doornbos warned that despite the apparent need for quick action and ready solutions in complex emergencies, there is a danger of reductionism. The scope for developing general policy responses to 'complex emergencies' and the food security problems posed by them remains very limited: better to start by assuming that each complex emergency requires its own analysis and response. This means less attention to searching for ready-made solutions, and more towards enhancement of institutional and operational capacities to diagnose emergency situations when they arise and as they develop. He also warned against automatic assumption of the link between complex emergencies and food insecurity. There is general acceptance that a food security crisis may develop as a result of failure to address other insecurities environmental, land tenure, livelihood as well as political but Doornbos reminded participants that complex emergencies may not involve
4

UN Inter-Agency Standing Committee Xth Meeting, 9-Dec-94

11

food insecurity (e.g. in parts of Somalia in recent years), or may be caused by food insecurity, for example when subsistence failures generate displacement. This highlights the need to avoid knee-jerk responses, in particular the mass injection of food aid. Doornbos concluded that because prediction of either the paths complex emergencies may take or restoration of the previous order is highly problematic, the key need is for anticipatory responses enabled by an analytical approach which is operationally relevant, light-footed, and flexible and not limited to food security dimensions. This implies a capacity for the kind of quick diagnoses which may be provided by a loose network of existing organizations with ears on the ground at macro or micro levels. In this respect agency comparative advantage is double-edged: it may privilege certain kinds of analysis and certain kinds of solution, hence the need for scrutiny of global institutional mandates and coordination of expertise. What is essential is effective analysis-action linkage: swift networking facilities to bring research and operations together so that actions reflect what is needed not what is available.5 Principles and methodologies in analysing complex emergencies Lionel Cliffe, Emeritus Professor of African Political Science at the University of Leeds, concurred with Doornbos on the very limited scope that complex emergencies offer for developing generalized policy responses, but sought to add two important observations, one on the ultimate objectives of policy, the other on methodology. First, he argued that the mode of analysis will necessarily depend on the principle objective of policy: is this to end the conflict, or to provide humanitarian assistance regardless of its effect on the conflict? Ongoing debates about humanitarian aims, e.g. impartiality versus do no harm, needs versus rights, have tended to sideline strong voices of the recent past arguing for the primacy of peace. While consequentialist in character rather than based on moral absolutes, these arguments are part of a debate that should continue to occupy practitioners and not just moral theorists, and take into account empirical findings about outcomes. Their logic is that food security initiatives should be guided not just by food needs assessments but also by security assessment and the needs of peace-building, though avoiding unintended outcomes will depend on understanding the political economy, as the experience of Operation Restore Hope in Somalia demonstrates (Box 1). This does however beg the question of who coordinates the analysis and the intervention. In several recent cases, most notably Iraq, the UN has been relegated to a secondary, humanitarian role by the intervention of outside states. A key analytical variable is thus the respective nature of military and humanitarian interveners and the relationship between them. Secondly, some agenda and methodology is needed to analyse complex emergencies, grounded in political as well as humanitarian realities. Cliffe offered a provisional analytical framework which examines four dimensions the state, the society, the dynamics of the complex emergency, and interventions and their consequences and four linked themes based on a) how delegitimization of the state generates violent conflict; b) how conflicts in turn reshape politics, the economy, institutions and the state; c) political and institutional legacies of conflicts and their implications for peace-building and reconstruction; and d) the need for a strategic, bottom-up approach to peace building, reconstruction and state reconstitution. As Cliffes paper demonstrates for Eritrea (Box 2), planning of relief and development also needs to recognize spatially as well as temporally differentiated war impacts and dynamics, shaped by shifting strategies of warring parties. The task of geographically disaggregated political analysis is to identify what kinds of conflict situations allow responses beyond basic relief.
5

The workshop paper by Gnter Hemrich, Matching Food Security Analysis to Context: the experience of the Somalia Food Security Assessment Unit, describes the forum approach to analysis-action linkage promoted by FSAU in Somalia.

12

Box 1. Food aid as instrument of peace in Somalia

In Operation Restore Hope (1992-3) the USAID-OFDA proposal to flood the country with food to reduce price and so remove incentives for looting of relief supplies and undermine warlords resource base reversed the original ORH aim of providing security for relief distribution The actual effect was the opposite of what was intended: the lower price meant more food was stolen to make up for its diminished value. Other unintended outcomes included adverse livelihood impacts on Juba valley agriculturalists and on incentives for grain merchants. As then OFDA Somalia coordinator Andrew Naitsos conceded, perhaps the two objectives of decreasing violence and increasing nutrition were mutually exclusive in the absence of a disciplined security force. A more informed understanding of the political economy of post-state Somalia would have made these effects predictable. Yet this does not invalidate all attempts to tailor food security intervention to conflict management or resolution objectives. Neglect of this dimension can effectively hand control to combatants who use relief for their own military/political objectives. Source: Cliffe

Embedding vulnerability in livelihoods models Such a broadened view of vulnerability in conflict-affected contexts was also advocated by Sue Lautze, Director of the Livelihoods Initiatives Program at the Feinstein International Famine Center, Tufts University. Presenting a paper prepared with her Tufts colleague Angela Raven-Roberts, she paid particular attention to its role in livelihoods models. People Box 2. Food security impacts of conflict in Eritreas liberation war
Studies in Eritrea by Cliffe and others at Leeds University in 1987, 1991, 1995 and 1997-99 explored the livelihood and food security impacts of war and drought in the period to 1991 and the extent of recovery during the interwar years 1991-98. These indicated a cumulative process of impoverishment and adaptation of production patterns and livelihood strategies in the war years. Adaptation meant loss of livestock and other productive assets and diminished long-term food security. One effect of the long conflict was a lack of agricultural and other statistics. The 1987 survey of 400 villages identified and mapped major agroecological zones: agricultural: southern and central highlands, with some 60% of population agropastoral: eastern and western mid/lowlands, with 30% of population pastoral: northern coastal plain and highlands, north-western lowlands and southern coastal region, with less than 10% of population War impacts in all these zones were profound, and reflected geographic incorporation in the conflict as well as production systems: Shifting front lines and fighting around them cut people off from crops and grazing lands, and sometimes displaced communities. In liberated areas, stated Ethiopian targets were rebel bases, but attacks also included villages and cultivated land and restricted movement, trade and food imports to night time (including the cross-border food relief operations via Sudan); In areas more under Ethiopian control, movement was circumscribed and young men and women brutalized as suspected rebels. Guerrilla activity provoked firefights, reprisals against civilians and detentions. Barriers between liberated and contested or occupied (including urban) areas fragmented the economy. Trade from northern and western liberated areas was oriented more to Sudan. Work in the mid-1990s sought to gauge how far this process of impoverishment had been reversed by the advent of peace in 1991 and what degree of recovery was in prospect. It highlighted: the multiplicity of livelihood strategies and levels of entitlement to food and productive resources within communities as well as across regions, though own food production remains central; that despite some improved harvests most households experienced continued depletion of oxen and herds, as well as an ongoing shortage of labour due to demographic effects of the war; that support to asset recovery would be a key element in rebuilding food security and livelihoods. Source: Cliffe

13

caught up in complex emergencies rely on their livelihood systems for survival, resilience and recovery, and recognition of the need to support these systems against collapse is not new. But beyond routine seeds and tools provision such support has not become a mainstream humanitarian response. Reasons include concerns about neutrality, donor unwillingness, difficulties of working in violent contexts and lack of understanding about livelihood systems and how they are affected by and respond to violence. While the DFID-supported Sustainable Livelihoods Framework (SLF) has been very successful in promoting a livelihoods approach in recovery and development contexts, its identification with the sustainable development agenda for which it was originally designed has limited the adoption of livelihoods-oriented programming by humanitarians. To be readily applicable in emergencies especially in violent, unstable and protracted settings the SLF needs to be reconceptualized with a focus on vulnerability and resilience rather than sustainability. In particular it needs to capture the dynamics of violence deliberately targeted at civilians and their livelihood systems. Provisionally, Lautze and Raven-Roberts propose a number of adaptations that would help in this respect (Box 3). First, the SLF depicts the vulnerability context as exogenous to livelihoods systems and only weakly influenced by processes, institutions and policies, whereas research suggests the components and dynamics of violence and vulnerability are endogenous to each component of the livelihoods model. Second, violence can transform livelihoods assets into life-threatening liabilities or, for the few, liabilities into assets. The assets pentagon needs to become an assets/liabilities pentagon to reflect this paradox for each type of capital (human, natural, financial, physical, social). Box 3: A humanitarian livelihoods model adapted for complex emergencies
PROCESSES, INSTITUTIONS & POLICIES

ASSETS/ LIABILITIES

INFLUENCE & ACCESS

STRATEGIES

OUTCOMES

FEED BACK FACTOR

Source: Lautze & Raven-Roberts

Third, processes, institutions and policies (PIPs) enable or hinder livelihood strategies, thereby generating or reducing vulnerabilities. Institutions, whether formal (e.g. government ministries) or informal (e.g. gender or age-based division of labour) can be vulnerable in complex emergencies. The policy environment can be a source of both resilience and vulnerability for households over time. International policies, religious extremism, health crises (e.g. HIV/AIDS), militarization and globalization also affect households ability to access and utilize assets. Research on violence offers lessons for humanitarians. Organized violence has deeper social as well as more obvious material impacts, distorting or replacing normal PIPs, e.g. by undermining existing networks of reciprocity based on trust and forming new ones based on fear, humiliation or torture. It lives on through generations, shaping future socio-political identities and blocking healing processes. Identifying practical implications of this is critical for enhancing livelihood interventions in complex emergencies. Finally, in the SLF livelihood strategies lead to positive livelihood outcomes including improved food security. Yet in complex emergencies actual outcomes are more often negative (starvation, poor health, mortality, destitution, shame, displacement etc.), and it is through the lens of societys PIPs that these outcomes are translated into assets but more frequently liabilities for households. As a basis for assessment and action, these outcomes are best defined by affected populations.

14

Discussion Participants discussed defining features of complex emergencies, some feeling they need not necessarily include violent conflict. For example Malawi in 2002 experienced a multidimensional and very complex emergency involving drought, floods, governance problems, HIV/AIDS and macro-economic crisis but not widespread violent conflict. The onslaught of HIV/AIDS has itself sometimes been described as non-armed violence. Crises such as the recent Asian financial crisis might also be considered in this category, and the role of incompetence leading to such crises should not be discounted. Others felt that crises involving widespread political violence are qualitatively different in that violence transforms all around it, including the nature of livelihoods and the state and institutions that would otherwise be partners for aid programmes, as well as restricting information and humanitarian access. The scale and types of responses that are possible become constrained by security risks, and interventions tend to be drawn into the dynamics of conflict and can even fuel it. There was general agreement that trying to arrive at a commonly agreed definition of complex emergencies is a semantic cul-de-sac, but this does not stop us addressing the issues. A preoccupation with root causes is also often unproductive, as conflict is in the nature of the human condition and original causes may fade in significance as effects become new causes. Whatever the causes, we are concerned with analysing consequences, mitigating impacts and increasing the resilience of those affected. In this regard there is a need for semantic clarity about concepts such as the role and fate of the state. Yet there is a danger of focusing on the effects of conflict rather than the conflict itself, which demands analysis if interventions especially those aimed at promoting peace are to have desired outcomes. It is important to bear in mind that for some actors conflict has a positive value including the means to change the status quo. Complex emergencies, like other disasters, also bring opportunities for positive change. Participants identified several key points emerging from the three panels: The field perspective

Vicious cycles between food insecurity and conflict, and the difficulty of action when
conflict becomes a food insecurity coping strategy: there is a need to address both symptoms and causes.

Apparent international indifference to humanitarian needs in remote or frozen conflicts,

and a failure to appreciate broader, historical contexts including international dimensions, conflict dynamics and local realities, and the disconnect between donor timeframes and longer-term societal needs in protracted conflict

The political economy of aid, the CNN factor and the large number of unaccountable
humanitarian agencies are not conducive to principled interventions.

The need to understand and address the role of gatekeepers, and the problems of trying
to achieve participation in a war situation: communities and field practitioners face similar constraints in this regard, but assisted communities should have a major input in decisions on how to assist and be dignified.

The problem of what to do about the political factors that triggered the complex
emergency, and the need to look for incentives for peace. The international agency perspective

Different agencies have different agendas and analyses, but there is a need to strive for
analytical consensus based on credible information, and a common approach to achieving a coherent response. However coherence means different things to different people. Is a commonly agreed analysis necessarily better than an awareness of the range of possible/different interpretations?
15

 Specificities of context, and especially the political ones, should not be overlooked in the
search for a common front: there needs to be a deeper understanding and less reliance on over-simplified, short-term technical solutions. Are we concerned with complex emergencies or complex food security crises? Political/structural causes of food security seem to be neglected by donors.

Tools for intervention are needed which go beyond humanitarianism. Humanitarian

assistance has at best little impact on peace building. But while enhancing analysis there is a need to accept the inherent limits of interventions. What kind of investments are possible in risky environments?

Large political actors tend to set the agenda, but UN agencies and NGOs also need to
develop common analytical tools. The research perspective

Planning response cannot avoid recognizing and addressing conflict, yet there is a need
better to understand violence and the related concept of powerlessness, and its links with livelihoods:
Response actions cannot avoid recognising/addressing conflict dimensions Mainstreaming violence and political economy analysis into livelihoods frameworks is

essential, but the livelihoods approach alone is inadequate to reflect the impact of conflict/violence
Violence can turn assets into liabilities Violence wins over institutions and leads to fear: vulnerability to violence needs to be

integrated into all aspects of the livelihoods model

Violence is also a means of livelihoods protection

Role of research
Is the research community sufficiently independent to challenge donors and

humanitarian actors?
Research can help avoid moving too quickly in identifying solutions to complex

emergencies by providing essential background information and analysis: a prerequisite for effective response
Researchers can often gain access to vulnerable groups without interference of

gatekeepers
Research needs multidisciplinary expertise to define the key questions for specific

contexts, but often lacks appropriate analytical tools

Research needs to provide complex analysis in simple sound bites so that it gets used

by practitioners: but there is a danger of simplified approaches as well as a need for simplified approaches, and an issue of how to use researchers analysis
Influencing conflict should be a prime goal of research

Session 2: Complex emergency processes, challenges and programming constraints experiences in different contexts
Introduction Session 2 aimed to compare food security impacts and programming challenges in a number of specific complex emergency contexts. Participants split into working groups covering major geographical regions:

Middle East/Central Asia Great Lakes

16

 Horn of Africa (Somalia/Somaliland) Horn of Africa (Sudan, Ethiopia, Eritrea) Southern Africa
Groups were invited to consider key contextual variables for specific cases, their food security impacts, responses of affected groups to these impacts, and constraints to and opportunities for longer-term food security programming to address these impacts. Middle East/Central Asia Group This group focussed on Afghanistan, and emphasized the role of external forces in shaping the present situation in which large numbers of people remain acutely food insecure despite a supposed post-conflict recovery. US foreign policy in particular has been the dominant influence in recent years, and is a strong determinant of policy for food security. Group findings are summarized in Box 4. Great Lakes Group The focus of this group was on the Democratic Republic of the Congo (DRC) and on the paper presented by Johan Pottier, Head of Department of Anthropology and Sociology, School of Oriental & African Studies. The war in Ituri in north-east DR Congo has killed 5060 000 people over the last 5 years, most of them civilians. Characterized as Lendu-Hema ethnic clashes, the conflict centres on access to land and its rich resources gold, timber, diamonds and coltan over which neighbouring Uganda and Rwanda vie for control. It also has important food security dimensions. Box 5 summarizes the findings of the group (in the summer, 2003) in term of key contextual variables and constraints and opportunities for longer-term food security interventions. Those involved in peace-building and agricultural rehabilitation must examine the social dynamic that gives warlords their iron grip on the population. They need to look at land, institutionalized vulnerability, the resulting need for institutionalized protection, and labour. These processes in Ituri follow a pattern rooted in social and land relations in colonial and post-colonial periods. The challenge is to plan for the removal of the conditions of insecurity that give warlords their grip on so-called ethnic followers. In addition to measures to protect and stimulate the postconflict resumption of local food markets, a commitment by FAO and other agencies to land reform in Ituri would help to reverse the regions extremely high levels of livelihood insecurity and thus weaken this stranglehold. Horn of Africa Group 1 (Somalia/Somaliland) The first Horn of Africa group examined Somalia, which featured strongly among the discussions at the workshop and was the focus of four of the workshop papers. The paper contributed by Gnter Hemrich, Food Systems Analyst in FAOs Agricultural and Development Economics Division, sets out the basic parameters of the complex emergency situation in Somalia as a prelude to an analysis of the FAO-run Food Security Assessment Unit. Since 1991 Somalia has been characterized by absence of government, a widespread but uneven and fluctuating pattern of clan-based conflict partly sponsored by other states in the region, self-declared autonomy of the relatively peaceful northern regions of Somaliland and Puntland but strained relations between them, a history of failed international initiatives to impose order by force and sponsor peace conferences, a collapse of systems for managing common property resources, and a periodic ban by Gulf states on Somalias main export livestock due to disease fears.

17

Box 4: Context assessment Afghanistan

Key contextual variables 18 years of conflict overlaid by a 3-year drought (1999-2001) A wide geographical diversity Religion and politics strongly intertwined, a strong sense of Afghan identity, and gender separation and exclusion. War not aimed at changing social structures, but external especially US interests have created/fuelled conflicts and affected societal structures, including the religion-politics mix. Influence of political, economic and intelligence structures of the US, especially in the last 2 years, and a legacy of US intervention High child mortality rates and personal insecurity A hidden economy (illegal and informal) has emerged to replace the legal/formal one. Access to land restricted and insecurity of tenure a potential source of political and economic conflict. Pastoralist livelihood systems destroyed. Large numbers of returning IDPs and refugees with few assets to rebuild livelihoods. Acute shortages of water due to destruction of water systems. Governance problems: collapse of state institutions, a weak concept of state and political instability. Funds to rebuild institutions are scarce, the concept of rehabilitation missing. Most donor support concentrated on the Ministry of Agriculture. Information systems are weak. Food security implications Availability is a problem primarily when there is a drought. The picture in the provinces remains unclear, especially in the south. With this years increased harvest, people are producing for local markets as well as own consumption. Access is linked to political stability, linked in turn to ethnicity, and there are problems of quality with a lack of dietary diversity especially in mountainous areas. Access is strongly gendered, with men controlling irrigation, harvesting and processing. Stability is affected mainly by high level of political instability as well as weather variability. PostTaliban, household incomes have become more dependent on the opium poppy market, which is subject to price fluctuation. Those without land are especially prone to instability of food access. Responses Short-term humanitarian interventions dominated by massive food aid and food for work A flood of ill-trained and ill-prepared international workers Efforts to establish an emergency unit Many projects and many kinds of food security interventions funded including seeds and cash

Constraints High levels of insecurity, limiting scope for participatory methodologies Government and institutional capacities unable to cope with interventions Powerful actors (CIA/FBI etc.) not technically prepared Limited knowledge of farm production systems, and limited use of what little knowledge exists Intense competition among agencies, but lack of ability among international staff and high turnover mean there is a response gap Donor influence sometimes perverse Opportunities Forthcoming elections Resilience of communities Substantial external resources available Gender is on the agenda The rebuilding of institutions offers an opportunity to redesign them.

18

Box 5: Context assessment Ituri, Congo DRC NB: Refers to conditions pertaining in summer 2003
Key contextual variables Huge mineral wealth controlled by the Ugandan and Rwandan armies and the respective rebel groups they support, with most miners forced into unpaid labour. The Ugandan army (UPDF) is also involved in illegal timber extraction and cattle raids. Administrative and economic dominance of the Hema minority over the Lendu majority. Gegere Hema in northern Ituri enjoy close trade links with UPDF, and backed Thomas Lubangas rise to power in the Union of Congolese Patriots (UPC) rebel group to facilitate control over spoils of war. The emergency was sparked off by clashes in 1999 between Lendu agriculturalists and Hema landowners who had bought ancestral Lendu land under the 1973 Bakajika land law. This law revoked automatic rights to ancestral land and decreed all land as state-owned but open to private sale. Lendu farmers, who considered the land inalienable, were evicted by Hema elites with UPDF support. Complex patterns of shifting alliances often linked to competition over land. Hema-Lendu disputes are not new, but were stoked into war by a mix of economic spoils, political manoeuvrings of national rebel groups, foreign backing by neighbouring states and international economic interests. UPDF strategy is to destabilize through exploiting ethnic sentiment for commercial gain. Vested interests in insecurity: warlords control over people and militias is based on control over land and threat of eviction thus Lubanga claims a cow, money or a child from each family. HIV/AIDS exacerbates peoples powerlessness. Women, responsible for 80 percent of food production, are disproportionately affected. Constraints Unwillingness to end conflict Private financial interests national, regional and international Failure to look into original causes Opportunities International pressure for land reform would help to reverse high levels of livelihood insecurity and thus weaken warlords stranglehold. Peace processes and initiatives, including measures to protect and stimulate resumption of local food markets. Further research on systems of institutionalized vulnerability and protection in Ituri Policy reforms, e.g. on land tenure, mining and access to pastures, to improve livelihood security.

Given all this, private sector growth in primary production, remittance and infrastructure services, trade, commerce and transport has been impressive, but is limited by lack of financial institutions. Remittances from the Somali diaspora provide the bulk of foreign exchange, but livelihoods and food security remain precarious for most, especially during the ban on livestock exports to the Gulf. Some 750 000 are chronically vulnerable of which half are displaced, access to rural services is very limited, primary school enrolment and adult literacy both stand at just 17 percent, under 5 mortality at 224 per 1000. The changing face of food security in Somalia Nisar Majid, Food Security and Livelihoods Adviser (East and Central Africa) for Save the Children (UK), noted in his paper that the period since 1991 has seen the nature of conflict in southern Somalia from inter-clan conflict centred around access to and control over particular resources and geographic areas to, over recent years, localized, intra-clan conflict. Most areas can today be described as generally unstable but with periods of stability. With the decline in external recognition and aid resources that followed the failure of UNOSOM, many of the major militia/faction leaders have either seen their influence diminish, or have changed the way they work. The formation of the Somalia Aid Coordination Body (SACB) in 1995 represented a profound change in aid strategy. This voluntary body includes NGO, UN and donor agencies and

19

facilitates information sharing, co-ordinates policy formulation and is designed to avoid the mistakes of large-scale, top-down aid programmes. An initial guiding principle was the notion of a relief-development continuum, in line with which Somalia was divided into zones of crisis (the south), transition (centre and north-east) and recovery (Somaliland) where security conditions were respectively deemed appropriate for relief assistance, rehabilitation support and development interventions. The continuum principle has since been modified in the face of criticism of its assumption of a linear progression from crisis to recovery, and in Somalia less deterministic descriptions of different parts of the country have been adopted, yet the principle of supporting coordinated, integrated and developmental interventions continues. The EC role in Somalia As Majid pointed out, the EC has been at the centre of this process in Somalia, with major funding in the field of rural development and food security directed towards food aid (via WFP), the FAO-implemented Food Security Assessment Unit (FSAU), and micro-projects with a rehabilitation-development focus. Emergency food needs have centred on the south of the country, where looting, fighting and instability have combined with natural hazards to make Bay, Bakool and Gedo the most food insecure areas. The FSAU has played an important role in independent analysis of emergency food needs in Somalia. Micro-projects have since 1978 included:

agricultural rehabilitation, mainly of irrigation infrastructure: initially with a 1-2 year duration
these have become increasingly developmental, with participatory and cost-recovery aspects and a longer time span; and

development of alternative products/micro-businesses (e.g. processing and marketing of

camel milk, oils, incenses and apiary products). Micro-projects emphasized rural livelihoods support to exploit peace dividends in the north and north-west especially and, though ad hoc in nature until the EC Somalia Country Strategy of 2002-2007, represented a commitment to longer-term intervention in Somalia unmatched by other donors. The EC Somalia Unit has largely managed to retain a high calibre and continuity of staff, a major factor in successful programming of longer-term interventions in such a constrained operating environment. It has also secured the agreement of its Brussels head office to progressively longer-term funding. Save the Children in Belet Weyn, southern Somalia Majid described Save the Childrens only remaining project area in southern Somalia in Belet Weyn in Hiran, arguably the most stable region in the south, though not at peace. Staff represent all clans and are influential at the elders level, avoiding the destabilization caused by other agency interventions. Links with local NGOs and CBOs have been avoided, as staff suspect their motivations and credibility. After several years distributing free seeds and tools, concern that this was undermining local coping strategies led in 1997 to an Agricultural Support Project (ASP) which as an ECfunded micro-project promoted local seed multiplication by sets of 3-4 villages, alongside agricultural extension, canal rehabilitation and later small water pumps. A perception by SC staff that the FSAUs focus on country-wide emergency food needs assessments was of limited practical relevance to NGOs concerned with more local, longerterm, non-food interventions led to a new emphasis in the Phase III ASP. A communitybased information system is being developed on the basis of existing community mechanisms for information collection and sharing under conditions of minimal literacy, along with development of seed fairs, water pumps, associated training and flexible livelihood support based on analyses by the information system. A key to success has been investment in national staff, who were formerly more familiar with emergency than participatory or developmental programming, along with continuity and
20

development backgrounds of expatriate staff. Larger NGOs have advantages in staffing across the emergency-development spectrum, and a stronger interest in informationresponse linkages. Ability to mobilize quick-release funds, e.g. for a seed voucher scheme after plantings failed, has also boosted the ASCs popularity. The ICRC in Somalia For the ICRC, Barbara Boyle Saidi (Head of Sector/Horn of Africa, Economic Security) and Christian Bosson (Economic Security Coordinator for Somalia) assessed the ICRCs strategy for working in southern Somalia which is two-pronged:

providing emergency relief assistance to mitigate the impact of violence and climatic
hazards;

restoring and protecting basic and productive (infra)structures benefiting vulnerable

communities, or located in areas where coping mechanisms have been weakened by debilitating crises. The ICRC recognizes that a sound understanding of the socio-economic political context is required to detect areas and communities potentially at risk, and to define the best way to strengthen existing coping mechanisms. In Somalias protracted but uneven conditions of conflict the ICRC has gradually shifted its focus from emergency relief towards rehabilitation projects. Targeted for ICRC intervention are specific communities made vulnerable by loss of productive assets land, livestock, wells, berkets (cement-lined water-storage pits), boreholes, tools, cereal reserves, irrigation facilities whose normal coping mechanisms (particularly social/clan support and seasonal migration) thus weakened or unavailable. The ICRC provides four types of assistance for restoring livelihoods of specific vulnerable groups:

Pump irrigation projects: rehabilitation of existing pump facilities along the upper and
middle Shabelle and Juba Rivers, plus a revolving fund for operation, maintenance and replacement.

Gravity irrigation projects aimed at flood control, water provision for agropastoralists, crop
diversification and land conservation along the lower Juba and lower Shabelle Rivers.

Community

intervention projects involving cash-for-work, maintenance of rainwater collection pools and berkets also fishing materials for riverine communities.

e.g.

rehabilitation

or

Seed distributions plus tools for specific communities unable to access or afford them;
The irrigation projects have been very successful. In 2003, 800-1 000 farm families with pump irrigation increased crop yield by 50 to 300% on around 1ha. per family, while 5 0008 000 families secured cultivation on 8 000 ha of farmland through gravity irrigation schemes. Pump owners, after initial problems of overuse, were encouraged to apply a system for tracking expenses, running hours, fuel costs, irrigated area, rainfall etc. Pump maintenance problems were addressed through a mobile ICRC workshop. Community-based projects benefited 5-6 000 families at little cost. Wage rates were a contentious issue: when set too high projects attracted the better-off to the exclusion of the most needy; when set too low projects were criticized as the ICRC seeking to exploit cheap labour. Local labour rates for weeding were eventually agreed on as a workable compromise. Also, communities did not initially identify with the projects they worked on. Reducing donated materials to a minimum, e.g. communities buying their own fuel and spare parts for pumps, helped to create a sense of ownership and accountability. In seeking to mitigate consequences for vulnerable people of shocks related to persistent armed conflict and recurrent climatic hazards, the ICRC is faced with the twin challenges of

21

Box 6: Context assessment Somalia

Contextual variables International, regional, national and local factors need consideration but no central government Heterogeneous contexts: different livelihoods, different conflict contexts: many unpredictable, volatile and dangerous, but some peaceful areas Also natural hazards, especially drought and floods. For many, access to food is marginal. Food security impacts include displacements, destitution, erosion of resources, a spectrum of high level (CNN) to low level impacts, and links between macro and micro impacts Responses International responses in certain areas mostly emergency, some developmental, some focussed on assets, either directly or through local partners Focussed on specific groups Some focussed on peace: problematic Partly directed by donor policy Level of consensus constrained by security perception, and question of what triggers action Constraints No state self-constructed systems of authority Perception by external actors: knowledge gap, labelling whole country as emergency Access (security, difficult logistics) Need to work from a distance, relying on local actors who are subject to influence of gatekeepers Weaknesses in generation, dissemination and use of knowledge Capacity building of local actors more trust Using areas and periods of peace Buy in gatekeepers Deepen the level of knowledge Support peace debates (state creation) Multi-scenario planning

Opportunities

addressing needs under unstable security conditions, and achieving project sustainability through involvement of beneficiary communities. The first is addressed by delegating responsibilities to national staff, but involves inevitable concessions to local power structures. The second necessitates a focus on rehabilitation strictly rather than building new structures, but also requires concerted attention to engendering a sense of ownership and accountability amongst those communities. Conclusions of the group are summarized in Box 6. Horn of Africa Group 2 (Sudan, Ethiopia, Eritrea) Based on work by Margie Buchanan-Smith, this group chose to concentrate on the 1998 Bahr-el-Ghazal famine in the Sudan as an example of an acute crisis within a chronic complex emergency. Group findings are summarized in Box 7. Another case briefly mentioned in the group discussions was that of the Eritrea-Ethiopia war of 1998-2000 and its food security dimensions. The paper by Philip White explores the food security impacts of the combined hazards of war and drought in the two countries, especially the 2000 food crisis in Ethiopia in the context of the war. It suggests that food insecurity and conflict were, and still remain, linked in more ways than generally acknowledged, and critically reviews the role of donors and humanitarian agencies operating in the two countries in this light.

22

Box 7: Context assessment: Bahr-el-Ghazal, Sudan in 1998

Contextual variables 1. An acute crisis within a chronic emergency 2. Pastoralist area: herd depletion after 3 years of drought led to increased vulnerability 3. Shifting militia allegiances: Kerabino, former SPLA commander, switched to Government of Sudan (GOS) side in 1992, but then attacked Wau in 1998 inducing mass flight to surrounding areas. 4. GOS responded with a flight ban to Wau, cutting off relief supplies 5. Background: aid diversion links to Operation Lifeline Sudan (Southern Sector) and SPLM conditionalities: show-me-the-dead; humanitarian but not development aid border trade and migration Food security impacts Massive mortality: 70,000 died Information systems picked up long-term deterioration but not militia developments: not much political analysis Aid: massive and costly inputs, and the use of aid to pursue political objectives Aid actors: greater emphasis on advocacy by external NGOs Donors found it difficult to use aid to change behaviours, but it can strengthen local actors Constraints Lack of policy framework in chronic emergency to gauge how well contingency planning is working and feed into policy Difference in level of inputs to GOS and SPLM areas Lack of political early warning and analysis Lack of linkage between policy frameworks (PRSP/MDG etc) and peace process Large groups of people concentrated in a few areas where food drops were permitted Opportunities Islands of tranquillity Lifting of flight ban UN contingency plan Peace negotiations

Lessons 1. The more complex the situation, the more open to manipulation it is 2. Political manipulation of humanitarian aid, by belligerents & external actors, delays response 3. Humanitarian aid alone is not a good tool for influencing a conflict and the behaviour of belligerents 4. Aid can change the relation between authorities, communities and other parties (POWs, crossborder operations) 5. Policy frameworks need to be tailored to context, not external blueprints (PRSPs etc) because of different capacities, access; general prescriptions around LRRD are very weak 6. External interests (oil, religious, War on Terror) have fuelled conflict 7. External actors have tried to influence the peace process

Volume 2 also includes the paper Harnessing the Potential of Aid to Protect Livelihoods and Promote Peace the Experience of the Nuba Mountains Programme Advancing Conflict Transformation (NMPACT), by Sara Pantuliano of UNDP Khartoum, whose participation in the workshop was prevented by mission duties in the Nuba Mountains. In contrast with Operation Lifeline Sudan, NMPACT represents the first attempt to bridge the longestablished division between Khartoum and Nairobi based agencies by bringing operations for the entire Nuba Mountains region under a unified management structure. It is designed as a phased, multi-agency, cross-conflict programme aimed at supporting all stakeholders to contribute to a Nuba-led response to peoples short and long term needs in the Nuba Mountains. NMPACT has a strong food security component, involving a political

23

humanitarian approach to addressing land tenure issues, underpinning advocacy action to return land illegally alienated to non-Nuba farmers. Advocacy has also been directed at addressing conflict between nomadic and farming groups and the harassment by the authorities of people returning to farms. Southern Africa Group The southern Africa group adopted a regional focus, but concentrated mainly on Malawi, where interactions between drought, floods and market and governance failures against a background of mounting poverty and spreading HIV/AIDS triggered the 2002 food security crisis, and Zimbabwe, where the government has blamed drought for a food crisis precipitated mainly by the politically motivated destruction of the commercial farming sector (Box 8). Box 8: Context assessment southern Africa, 2002
Contextual variables A continuum between contexts defined by political violence (Zimbabwe) and others (Malawi) HIV/AIDS as a contributing factor rather than a cause, straining coping mechanisms Liberalization undermined food security removed institutions, not replaced by market mechanism Underlying chronic vulnerability discovered/exposed in emergencies Zimbabwe: commercial agriculture deliberately destroyed as part of ZANU/PF bid to stay in power Infrastructure and transport networks good but problem of dealing with vulnerable pockets Low purchasing power access issues Market and safety net gaps

Food security impacts 1. Declining coping capacity 2. Access and stability affected at all levels, availability at household level 3. Stability issues: pronounced weakness in markets (Zimbabwe effect, liberalization) 4. Availability: negative impact of input/output terms of trade Opportunities 1. Facilitating environment (except Zimbabwe): no violent conflict, little politicization, mostly good rapport with donors, no problems of access to affected populations 2. Adjust liberalization policy to correct evident failures, especially balance between state and markets 3. Focused recognition of the food security implications of HIV/AIDS: search for alternative livelihoods with lower labour inputs; need for scaling up of safety nets Constraints 1. Food security not an interest of the present Zimbabwe government 2. Governance issues governments agendas, e.g. urban bias in Malawi 3. Weakness of institutions in new democracies 4. Weakness of regional institutions 5. Long timeline of turning round HIV/AIDS

Discussion The diversity, uniqueness and complexity of different contexts was noted. Politics is a prominent and recurring theme running through these cases, highlighting the need for political analysis as a basis for action to address food security impacts. Key contextual variables brought out by the groups were:
the role of the state, and failures or lack of governance political influences on aid policy (Afghanistan, Sudan) HIV/AIDS (southern Africa) the political role of external actors (Somalia, DRC) 24

 very different contexts in different areas (Somalia) conflict over natural resources, exacerbated by foreign interests (DRC).

Participants felt that the drive on the part of donors to push towards rehabilitation and development sometimes more apparent as a drive simply to wind down humanitarian assistance needs to involve better context-specific analysis if it is to take advantage of islands of peace in situations of chronic conflict and instability. An emergency does not necessarily equate to a need for relief developmental or livelihoods-oriented interventions, e.g. involving dam-building or establishing a blacksmiths cooperative, can be implemented on an emergency basis or during a protracted crisis. Despite the danger of generalizations, lessons can be learned which are applicable in other contexts including lessons from successes. One such case was offered for consideration: the cross-border operation to provide relief and development assistance in Eritrea during its liberation struggle this was an act of solidarity with a then rebel movement (EPLF) which not only achieved humanitarian access but boosted popular support for the EPLF. Furthermore, context-specificity does not necessarily mean different activities in each case: many of these can be common to different contexts. The role of the private sector should not be overlooked and opportunities for foreign direct investment to play a positive rather than negative role. The most appropriate balance between private and public sector activity is not a foregone conclusion, but itself depends on the context. Finally, it was noted that the case studies demonstrated ways in which food insecurity is not only a by-product of violent conflict but is often also a weapon and a deliberate war objective.

25

Theme 2: Principles, approaches and policies for food security in complex emergencies (Session 3)
Introduction
Session 3 addressed the second main workshop theme: how best to move from contextual analysis to appropriate responses for supporting the resilience of food systems in complex emergencies. Plenary presentations were made by a panel of four speakers: two involved in applied research, and two FAO staff responsible for supporting policy formulation and implementation. This was followed by group work to consider the formulation of policies in contexts with different degrees of risk of violent conflict:

Plenary panel
Back to basics: the importance of humanitarian principles Margie Buchanan-Smith, an independent consultant, sounded a note of caution on the importance of observing key humanitarian principles in responding to complex emergencies. Their importance derives partly from the practical need to gain access to affected groups: if agencies do not rigorously strive to be neutral, impartial and independent they are in danger of being seen by one party to the conflict to be favouring another, and their access will be restricted. Do humanitarian principles contribute to the progression from contextual understanding to longer term policies for food security? No and neither should they. Contextual understanding is essential, but this should not be about longer term policies. There has been a move (exemplified in some contributions to this workshop) to adopt a broadened concept of humanitarianism which includes developmental interventions aimed at livelihoods and peacebuilding ones aimed at modifying conflict and changing behaviour. The effect of this is to confuse aims and dilute key humanitarian principles which demand that humanitarian assistance is provided unconditionally i.e. not contingent upon its ability to achieve development or peace. The result of this broadening is that negotiations with belligerents are hampered because they perceive ulterior motives which may not be to their advantage. Buchanan-Smith argued for a minimalist approach in humanitarian assistance which is less likely to interfere with the dynamics of conflict. She nevertheless recognized the need to complement humanitarian action with longer term interventions. This could be achieved by one or two agencies specialising in pure humanitarian assistance, freeing up other agencies to pursue other longer-term objectives such as livelihood interventions. HIV/AIDS and Food Security: a long term emergency6 Marcela Villarreal, Chief of FAOs Population & Development Service (SDWP), gave an account of links between HIV/AIDS and food security and the longer term prognosis as the epidemic runs its course in countries with high infection rates. HIV/AIDS impacts on food security in a number of ways including loss of agricultural labour, impoverishment of households, loss of knowledge and the macro-economic impact of
6

See also the paper prepared for the workshop by Michael Loevinsohn and Stuart Gillespie (whose travel to the workshop was prevented by Hurricane Isabel). Entitled HIV/AIDS and Food Crises in Southern Africa: an agenda for action research and for learning how to respond, the paper sets out challenges to addressing HIV/AIDSfood security interactions in the form of gaps in understanding, limited capacity to act and lack of evidence on what works. It presents the HIV/AIDS lens as a conceptual tool for decision-makers, outlines the work of the Regional Network on HIV/AIDS, Rural Livelihoods and Food Security (RENEWAL) in Africa, and discusses action research priorities involving partnerships between development and humanitarian thinkers. (See Volume 2)

26

Low Prevalence High Impact High Prevalence High Impact High Prevalence Low Impact Low Prevalence Low Impact

Prevalence

Deaths

Double Orphans

Figure 3: HIV/AIDS: trajectories of prevalence and impact increased food imports. Projected losses in the labour force due to HIV/AIDS in the most seriously affected countries of southern and east Africa during the period 1985-2020 range from 13% in Tanzania to 26% in Namibia. Households are impoverished through having to sell of productive resources for care/funerals (e.g. in Ethiopia average treatment and mourning costs exceed average annual farm incomes; while in Kenya 49-78% of household income is lost when one person dies from AIDS), and a shift to less labour-intensive crops, which means less cash crops and less nutritious crops. The sustainability of food production is undermined through:

undermining of transmission of knowledge between generations (household and

institutional)

a decrease in range of crops leads to decrease in plant diversity and genetic resources social safety nets are undermined, leading to an irreversible collapse of the social asset
base. Risk factors associated with HIV/AIDS include migration, undiversified sources of livelihoods, gender based violence, gender inequality and food insecurity. Macroeconomic implications are profound. Average annual loss in GDP growth per capita due to HIV/AIDS is estimated by the World Bank at 0.7% in Africa (0.6% in Ethiopia, 1.3% in Kenya, and 0.8% in Uganda). Figure 3 shows three hypothetical curves respectively representing prevalence of HIV infection, and impacts due to AIDS deaths and to the number of orphans who have lost both parents to AIDS. Assuming infection rates can be brought under control, the peaks for these three variables can be expected to be staggered over time approximately as shown. Four phases can therefore be distinguished: an initial phase of low prevalence and low impact, a second phase of rising prevalence but still low impact, a third phase in which prevalence and impact are both high, and a fourth phase when prevalence has declined but impact is still high. The present situation in southern and east Africa corresponds to the early part of the third phase in other words even when measures succeed in reducing prevalence, impacts and their food security implications will continue to intensify for some time. Villarreal concluded by considering responses in this light. A balance between immediate and long term responses must be struck, but emergency interventions with a development perspective or interventions that take into account the long term are required. There is still a

27

window of opportunity to take action to prevent the worst impacts of HIV/AIDS on food security. FAOs Emergency and Rehabilitation Operations: an evolving institutional perspective Richard China outlined the way that FAOs approach to and response to emergencies is evolving in response to the challenges already discussed in the workshop. FAOs traditional role (development-oriented with a small seeds/tools-focussed emergency programme) is changing; and donor funding for the emergency programme has grown substantially. A 2001 evaluation of FAOs emergency operations in the context of the crisis management cycle has prompted a number of changes:

institutional changes (within FAO and with partners) aimed at mainstreaming emergencies
and institutionalizing intervention strategies;

the establishment of a unit for linking emergency work to rehabilitation and development; a review of procedures for early warning and crop and food supply assessment; a shift from an input-led bias towards livelihoods-oriented interventions.
Progress in implementing these recommendations has included: a new Emergency Division and Policy Unit has been set up, with new posts, guidelines, and typology of interventions
seed aid has been reviewed, and new interventions such as seed vouchers designed; seed security assessments are made within a food security framework; policy on social protection and safety nets has been reviewed; interventions for strengthening self-reliance, and for involving the private sector have

been designed. The above is backed up by much new thinking on approaches to emergencies within FAO7. But there is still a need to overcome institutional (and individual) resistance to change and this is more difficult. Key questions that remain for FAO are:

How can FAO best promote and implement policy changes? Do good assessments/proposals really have an impact on (donor) resources? And how
can one ensure that resources allocated where they are needed most? Coming to terms with complexity Franois Grunewald, Director of Groupe Urgence Rehabilitation Developpement (Groupe URD), an umbrella group of some 30 French and European NGOs concerned with relief and development issues, considered the challenges of developing policy responses in complex emergencies. The burgeoning galaxy of aid The acute suffering of non-combatants, increased media attention and reversal of any development gains associated with these conflicts has led to a burgeoning of humanitarian aid. Numbers of agencies have multiplied, initially largely Western-based, making
7

This includes an ODI-FAO-ICRISAT research project on agricultural rehabilitation, the inception report of which is summarised for this workshop in the paper by Kate Longley with Ian Christoplos and Tom Slaymaker, From Relief to Food Security? The challenges of programming for agricultural rehabilitation (see Volume 2). The paper reviews recent innovations in seeds and tools programmes, market-based interventions, accountability issues, livelihoods oriented programming and institutional capacity-building, and sets out key challenges for agricultural rehabilitation in conflict situations.

28

coordination a key issue. These cover a wide spectrum of approaches and interests, some implementing donor policies, others espousing independence. Some NGOs have opposed prevailing concepts of international humanitarian law, only to rediscover them subsequently. More recently Southern and Islamic NGOs have proliferated too. UN agencies have engaged in unseemly turf wars which have undermined proper coordination and prioritization in emergency response and devalued the consolidated appeals process. External armed forces have also become involved in crisis management, in a wide variety of roles including peacekeeping, political nation building, policing, security and humanitarian provision, often with mandates that are far from clear. Private sector involvement, e.g. in logistical support, reconstruction and armed security, has also become a feature of international response. Aid actors and humanitarian workers are increasingly being deliberately targeted. Perpetrators feel they have little stake in international humanitarian law, and the integrated approach embedded in various forms of Strategic Framework and in the coherence agenda may not help in this respect. Afghanistan and Iraq are recent examples where the US doctrine of either for us or against us leaves NGOs little room for independence, neutrality or impartiality, three key principles of humanitarian action. The necessary diversity of responses Despite the homogenising discourse of exclusivist, consumerist thinking growing in the humanitarian field, it is important to take stock of the experience gathered in responding not to sector-based needs but to the holistic nature of human beings, strengthening resilience and capacity to implement coping strategies, with due attention to whose needs are being addressed. Flexibility and ability to plan for multiple scenarios are crucial in crisis situations, but are not well served by the bureaucratic structure of most donor agencies, which tends to emphasize project cycle management techniques used in stable development contexts and upward financial accountability rather than downward accountability to beneficiaries. Universal standards devised by international NGOs are reassuring, but are paternalistic in essence and may be at odds with priorities emerging from local knowledge. Who should identify relevant indicators for interventions? Challenges ahead The management of crisis needs to change because the nature of crises is changing. Challenges ahead include HIV/AIDS which undermines the basis for resilience, climate change, small arms proliferation, new weapons and communications technologies, the growth of informal and war economies, increasing religious and ethnic tensions and growing inequality. New methodologies are required, e.g. multi-scenario approaches, enhanced participation, an increased role for qualitative social analysis. Confronting complexity with intelligence and turbulence with flexibility are not easy tasks and there are no easy solutions. Discussion Participants raised several issues for the panel to consider:

Should FAO establish a think-tank on food security issues in complex emergencies? If

so, this should relate closely to operational aspects of complex emergencies, including food security information systems.

Who are the food insecure? They are not necessarily agricultural producers, and there is
often insufficient emphasis on urban areas where people lack income to buy food. A broader livelihoods approach is important.

Three key approaches under complex emergencies are humanitarian action, livelihoods
29

interventions and conflict reduction/prevention. Short-term humanitarian action is guided by a clear set of principles (humanity, neutrality, impartiality etc.) but what principles should guide longer-term livelihoods and conflict-reducing interventions in protracted crises?

Urbanization, war, AIDS, malaria and other factors all have had a negative impact on
labour. War-ravaged Eritrea has faced a problem of labour scarcity for some time, but also of draught animals this led to a mechanization policy, which has its own problems. In pastoralist societies, young people are moving into town, putting pressure on rural labour availability. A decline in the labour force as a result of HIV/AIDS is likely to lead to collapse in farming systems, in such a way that labour-saving technologies may not help. Rebuilding institutions will be important in dealing with this. Too little attention has been paid to the impact of the collapse of local institutions during complex emergencies this needs further research.

Malaria is as devastating as HIV/AIDS yet receives much less attention. In relation to

HIV/AIDS risk factors, governance issues should be included as the Malawi example shows. Panel responses highlighted the following points:

Think-tanks are most useful when they are close to operations. The UN Resident
Humanitarian Coordinator in Sudan chairs a cabinet of local aid actors who review policy and take decisions. In Afghanistan, there is no adequate information strategy to support policy. Womens empowerment can be an objective of emergency agricultural operations: getting a cow and poultry into peoples backyards in Afghanistan would further this objective, but actual implementation remains a problem.

In practice, most humanitarian assistance programmes have components of conflict

resolution and peace building as well as development: this sends confusing messages to recipients. This blurring of mandates compromises the ability to undertake humanitarian action. Other kinds of interventions have a place, but there is a need to consider how to protect humanitarian space.

Labour is the most basic resource, embodying energy, time and knowledge. Labour
shortages are/will be particularly acute in Africa, and drought-conflict interactions will exacerbate the problem. Coping mechanisms tend to be labour-intensive and so will become less viable.

Given the emphasis on HIV/AIDS impacts, policy implications of other key diseases such
as malaria should not be overlooked. Agreed: integrated responses are needed; but HIV/AIDS does present a special case given its complex impact on society. Whereas malaria attacks the very young and is not increasing much, HIV/AIDS kills the most productive and reproductive age groups who are also responsible as carers

How risk of violent conflict affects policy choices for supporting food security
As a way of focussing discussion on policy issues, groups considered the likely appropriateness of a range of policy/programming options in relation to different degrees of proximity or risk of violent conflict. This exercise was designed not as an attempt to make specific policy prescriptions for generic types of situation something that had already been warned against in the opening plenary session but to permit exploration of ways in which policy choices for food security need to reflect political and security conditions. Groups were organized around the following set of conditions:

Group 1: Multi-dimensional crises without violent conflict Group 2: Situations in which there is a threat of violent conflict Group 3: Acute crises involving violent conflict
30

 Group 4: Protracted crises when there are lulls in conflict of uncertain duration Group 5: Situations in which longer-lasting peace seems likely
Each group was invited to consider:
What kinds of interventions are likely to work in supporting food security (in terms of

availability, access, stability)?

What kinds of interventions are unlikely to work in supporting food security (in terms of

availability, access, stability)? What are the requirements for making interventions work?
What actors are involved?

The results of this exercise are discussed below. A summary matrix of policy options was generated which is included as Error! Reference source not found.. Group 1: Multidimensional crisis without violent conflict (Table 1) Group 1 pointed out that in the absence of threat of widespread violent conflict, constraints for longer-term support to food security are fewer and options wider. State and parastatal institutions are not compromised by association with or control by the state as a belligerent party, or by factions fighting for control of the state. Markets are able to function within a legal framework which is operational and has some legitimacy even if it is not always respected. Agencies can engage in humanitarian and development assistance without access to needy populations being subject to massive security constraints, and without having to face the dilemmas of whether their activities are neutral and impartial or complicit in human rights abuses or build peace or fuel war. Private sector, civil society and community actors are not involved in routine physical violence either as perpetrators or victims. The principle of linking relief, rehabilitation and development is not as contentious as it is in the politically-charged atmosphere of complex emergencies involving conflict. Malawi in 2002 was noted as an example of a crisis which was multidimensional in both cause and effect and involved a descent into famine conditions in several parts of the country, but did not feature widespread violent conflict. Key issues Several key issues arising in the Malawi case have wider relevance for others in this category:

Donors tended to pursue their own agendas without due regard to the overall coherence
of the crisis response effort or the longer term food security implications. Emergency responses to the crisis were themselves delayed by donor mistrust of the government, as had occurred in Ethiopia in 2000.

Even in the absence of conflict, social networks and the informal safety nets they can
sustain are subject to degradation when livelihoods become so precarious across whole communities that individual groups and households have little to spare to meet social obligations to help those in acute need. Interventions need to reflect an awareness that households are not isolated entities, and social networks need to be supported and rebuilt over time.

Targeting of support towards the most needy is a prominent issue. Alternative targeting
methods, all of which have their drawbacks as well as advantages, need careful consideration with respect to both coverage (avoiding exclusion errors) and efficiency (avoiding inclusion errors), set against their costs and timeliness. What works The group considered what works for each of the three food security dimensions:

31

Table 1: Multi-dimensional crises without violent conflict

Key issues Without conflict, constraints are fewer and options wider. Public and private institutions can play an important role. Donors tend to cherry-pick options to suit their own agendas, leading to possible problems of programme coherence. Strong links are needed between interventions, research and policy frameworks. Need to utilize/support/recreate social networks. Increased access is the aim but direct free distribution is risky (alters incentives) Cost-effectiveness of different targeting methods need to be examined. Availability: Low input/less labour-intensive technologies Measures to increase access to agricultural inputs, e.g. vouchers for agricultural fairs; inputs-for-work Improving access to services, and enabling private sector involvement Access: School feeding plus (increased enrolment, lower drop-out, higher achievements) Increasing effective demand, e.g. with cash/food/vouchers for (asset creating) work Measures to build up household assets Safety nets for orphans and the elderly Stability: Reduce physical and fiscal barriers to trade Support grain reserves (local, national, regional, grain futures) Innovative micro-credit schemes (?) What doesnt Availability: free agricultural inputs; business as usual for agricultural extension work Access: relying solely on the market General: over-complex projects that cannot be scaled up, replicated or sustained Accountability of all players, and effective learning processes Whats needed for Targeting and empowering the most vulnerable including self-targeting interventions Coordination: to maximize economic benefits of social protection; PRSPs? to work Better early warning indicators of access and availability; market information systems Appropriate incentives: economic; institutional Transparent management of grain reserves Research/understanding of context and dynamics Institutional capacity building; revisiting the role of institutions - new models Large scale outreach potential What actors are involved Institution Min. of Planning/Finance Parliament/MPs Executive/Cabinet Ministry of Agriculture National AIDS Council Farmers/Producers Associations Donors Relief/humanitarian/development agencies Research community Private foundations Private sector NGO/civil societies Role PRSP, Budget Legislative Decision makers Food production, standards, quality, early warning, monitoring Programming/coordination Lobbying, market access Money, prioritization Technical assistance Research, lessons learning Finance Service provision Advocacy, implementation support

What works

32

 Availability can be boosted by a range of measures to widen access to agricultural inputs,

including inputs for work and input vouchers. Where household labour is a constraint as is increasingly the case in many AIDS-affected rural areas support for technologies which are less demanding of labour is likely to be better received. Improved access to agricultural services, including via support for private sector service establishment, can relieve a major constraint where input supply and produce markets and credit systems do not function effectively, as in Malawi where private sector service providers have yet to fill the gap left by state-led services following liberalization measures.

The group noted the role of social safety nets in protecting food access, including workbased schemes for households with available labour which can create assets capable of boosting food security in the longer term, school-feeding programmes and alternative forms of transfer for orphans and the elderly.

Stability can be enhanced by measures to reduce trade barriers, and to support the
establishment and careful management (unlike in Malawi in 2002) of grain reserves at various levels. What doesnt work The group noted that the free distribution of inputs to large sections of the rural population, as in Malawis Targeted Inputs and later Starter Pack programmes, has the potential to boost overall food production substantially if season quality is favourable and if delivery is timely, but is fiscally unsustainable in the longer term and does little to assist the development of viable private sector marketing systems. Agricultural extension systems need to adapt to the needs of the most food insecure, who are most often unable to afford costly inputs, take risks or, where AIDS impacts are substantial, engage in labour intensive methods. Too often extension advice remains geared to small commercial farmers who are in a position to invest in yield-enhancing technology. The group also noted the risk associated with interventions which rely on properly functioning market mechanisms and infrastructure which are not actually in place or which exclude the majority of the poor and food insecure. Whats needed for interventions to work The group identified several pre-conditions for interventions to yield positive results. In particular, systems for effective accountability and learning from experience are essential, yet are generally at an early stage of development (a pilot project to establish a Learning Support Office has recently been completed in Malawi). Early warning and information systems need to improve their indicators of both availability and access in the Malawi case there was an overestimation of the availability of root crops in the period leading up to the 2002 crisis). Targeting mechanisms need to identify and reach the most vulnerable, and include selftargeting involving innovative work-based schemes (cash/food/inputs/vouchersfor-work) which create useful assets. For HIPC countries the PRSP process can fulfil a valuable function as a framework for national coordination of initiatives including donor-supported ones to improve food security, as well as other aspects of poverty reduction. What actors are involved The group found that a wide range of actors, as shown in Table 1, have a role to play in making effective longer-term interventions for food security, spanning government, nongovernmental, donor and civil society organizations. This highlights the challenge of establishing policy frameworks which can serve to harness the efforts of multiple actors to achieve greater coherence towards this end.

33

Group 2: When there is the threat of violent conflict (Table 2) In looking at contexts in which there is a threatened outbreak violent conflict, this group identified three broad types of situation: 1. Dappled conflict where there is a mosaic of conflict and peace with only limited though often shifting areas involved in violence, exemplified by Kenyas northern districts, Eritrea during its liberation war and Afghanistan under the Taliban; 2. Frozen conflict which could resume at any time, as applies to the disputed enclave of Nagorno Karabakh over which Armenia and Azerbaijan went to war in the late 1980s and are still technically at war despite a ceasefire in 1994; 3. Situations labelled post-conflict, such as contemporary Afghanistan and Cambodia, where an uncertain peace has the potential to descend again into renewed conflict. What works The group emphasized the importance of support for proper functioning of markets:

enhancing their resilience in the face of a range of disruptive influences including

speculation, misguided state intervention, inappropriate food aid and high value and perhaps illegal cash crops;

addressing

decapitalization/destocking as adaptation to risk by supporting recovery/reconstitution of productive assets (though short term answers can cause longer term problems); information.

enhancing marketing infrastructure: transport, sites and storage, communication and

Other interventions options include helping to make agricultural systems more productive and resilient, developing or supporting basic services where conditions allow, safety nets which can stem decapitalization, measures to improve targeting, build up institutional capacities, create employment, protect peoples access to natural resources. Food aid, where appropriate and timely, can play a positive role. Whats needed for interventions to work The group stressed the importance of improved early warning and information systems, employing cost-effective methodological tools to enable the timely identification and analysis of seemingly unrelated signals and their targeted dissemination. These signals cover many different dimensions, from food security to social/political tensions, economic degradation, lawlessness, media manipulation, market failures, access to natural resources, behaviour changes and proactive population movements. Such systems need to link with research to result in an adequate understanding of actual and potential conflict dynamics and impacts, of institutions and of informal safety net systems which can be supported. What actors are involved Another main focus of discussion for this group was the wide range of institutions and actors likely to be involved, including community/cultural, traditional, political parties, government and its internal dynamics and conflicts of interests, external actors whether NGOs, economic actors or ghost actors such as the CIA. The identity of these various actors, their roles, relations of conflict, their strengths and weaknesses in knowledge, communication and capacity are important determinants of what can be achieved. Opportunities for capacity building/rebuilding and for positive change in roles and set-up need to be understood and identified, and such change needs to happen before conflict erupts. These roles may relate to interventions in food markets, land tenure, access of the vulnerable to water and natural resources and the legislative framework for protecting such access and its enforcement (e.g. where there is invasion of pastoral land).
34

Table 2: Situations where conflict is threatened

3 types of situations 1. Dappled conflict: a mosaic of conflict and peace, e.g. Kenya, Eritrea, Afghanistan 2. Frozen conflict, e.g. Nagorno Karabakh 3. Post-conflict, e.g. Cambodia, Afghanistan

Early warning: Cost-effective methodological tools to enable identification and analysis of seemingly unrelated signals (food security, social, economic, demographic, political) and their timely, targeted dissemination Institutions & actors: (community/cultural/traditional, CBOs, political parties/government, external) and their roles, potential/actual conflicts, strengths and weaknesses, opportunities and threats Markets and private sector: resilience, decapitalization/destocking, infrastructure What works Support for proper functioning of markets, including local purchase of food/inputs Basic support to agricultural production and resilience of agricultural systems Establishing/supporting basic services - theatre of normality Actions to reduce destitution/decapitalization, including safety nets Targeting of interventions (versus blanket coverage) Reinforcing capacities of institutions Create employment - even temporary Involvement of diaspora Food aid where appropriate and timely Improve/protect access to natural resources What doesnt Involvement of diaspora work Knee-jerk rehabilitation interventions Reliance on markets only Illegitimate, badly managed coordination mechanisms Poor targeting (high exclusion/inclusion errors) Food aid where it undermines livelihoods Improve early warning and food security information, public information, advocacy Whats needed for Innovative methodological tool for timely measurement and analysis of indicators interventions Analysis of potential conflict impacts to work Involvement of local actors Develop legal instruments to improve access to resources Better understanding and support of social/informal safety nets Good institutional analysis Appropriate feasibility study (as participatory as possible) Disarmament, demobilization, reintegration (DDR) in post-conflict situations Networking and support of local NGOs and CSOs and improved coordination Disaster preparedness strategy/contingency planning Effective advocacy for resource mobilization for food security and political action What actors INGOs and Red Cross are involved Government UN agencies Winners and losers Ghost actors (including 'spooks') Other governments - regional and others Civil society - CSO, CBO, NGO, political, private sector International financial institutions National and international media Key issues

35

Group 3: Acute crisis involving violent conflict (Table 3) Key features Group 3 began by exploring likely attributes of contexts which fall into this category and observed that there may be more than one peak period of acuteness, and successive peaks may differ in important ways. There is a need for clarity about what acuteness actually entails. Most often physical insecurity and a need for humanitarian protection are key features, but immediate food insecurity may not be. Common features include high mortality and morbidity as a direct result of violence or through disease and war-induced starvation, heavy impacts related to displacement and large concentrations of displaced people. Civilians may be held hostage, and their livelihoods and assets deliberately destroyed. Infrastructure and institutions also tend to be destroyed, and normal markets and social networks breakdown or are replaced by ones based on coercion and violence as part of an illegal economy which may include the use of compulsory labour. Violence, discrimination and subordination are often based on ethnic or clan identity. However, as the case of Ituri, DRC illustrates, allegiances may shift both locally and regionally, transcending ethnic boundaries. Conflict may, as in DRC and in the Greater Horn, be driven by regional dynamics, though there are invariably national aggravating factors. Flows of information across borders tend to be very restricted, and there may deliberate campaigns of disinformation by militias and/or regional governments. The group identified a range of contexts which might fall into this acute crisis category (listed in Table 3) and noted that while many of these exhibit many of the above features, it is risky to attempt to design policy on the basis of any such generalization as each context is distinct with respect to a wide range of attributes (as described in the paper by Cliffe) which critically affect what works and what doesnt work. What works Accordingly one main conclusion about what works is: it depends ! The group noted that the overriding objective in such situations is to save life by emergency implementation of measures to protect civilians and deliver humanitarian relief. This may or may not necessitate food aid delivery to increase food availability and a range of measures to protect food access. Initiatives that may be relevant in acute crisis contexts include safe havens to protect civilians; market interventions to lower food prices, perhaps working through local merchants; advocacy and other efforts in the political arena (not necessarily by food security agencies) to promote dialogue, defuse conflict and establish incentives for warlords to allow access; negotiations for return of stolen cattle, linked to animal health programmes; the provision of inputs such as quick maturing vegetable seeds which can help those with land to generate cash in local markets as well as improve nutrition; where necessary food aid (even where it does risk being reclaimed) as well as therapeutic & supplementary feeding programmes. Cash injections, perhaps through work-based programmes, can be used to stimulate local markets. In general there is advantage in building on existing programmes and structures. The range of different distribution & targeting techniques needs to be considered. Whats needed for interventions to work Echoing other groups, this group emphasized the importance of coordination between interveners. Apart from anything else this increases the chances of being able to counter predation strategies effectively, as does sticking to interventions which are relatively low in profile. Agencies need to have well-differentiated roles with clear objectives and timeframes, and experienced staff capable of detailed contextual analysis, creativity and flexibility in programming and able to draw on adequate funding and resources. Other important conditions for success include making full use of existing information and knowledge, including what can be learnt from past interventions, including wide local

36

Table 3: Acute crises involving violent conflict

Key features Not necessarily one peak period, and peaks not identical What does acuteness entail? immediate food insecurity? Not all crises involve food insecurity physical insecurity and need for humanitarian protection Extreme mortality/morbidity/malnutrition Heavy displacement/migration/impacts, large population concentrations Civilians held hostage Deliberate destruction of livelihoods; breakdown of markets and social networks Physical/institutional destruction Illegal economy including compulsory labour Actors/militias shifting local/regional alliances Often a regional dimension, with national aggravating factors Lack of information across borders, creation of disinformation Ethnic identity/discrimination Fed. Rep. Yugoslavia, 1993/94 Cambodia, 1973-75 Kosovo, 1999 Biafra, 1968-69 East Timor, 1999 DPRK (?) Goma, 1994 Uganda/North, multiple years 1990s DRC, 1997 Liberia/Monrovia, 1996, 2003 Rwanda, 1994 Ituri, DRC, 1999 onward peak: militia fighting AprSep 03 Colombia South Sudan, 1983 onwards and in various Angola, 1992, 1997 parts of the country Afghanistan Somalia, 1991/92 Chechnya

Examples & time frames

It depends, but the overriding objective is life saving, including increasing food What interventions availability (food aid delivery) and access. Options may include: work? market interventions to lower food prices, and use of cash to induce markets working through local leaders/merchants for delivery quick maturing vegetable seeds - to promote via vegetable markets negotiations for return of stolen cattle & links to animal health food aid (even though its reclaimed), therapeutic & supplementary feeding dialogue/political processes - but not by food security agencies necessarily advocacy for political processes & incentives for warlords to allow access build on existing programmes & structures different distribution & targeting techniques safe havens NOT addressing root causes, but trying to do no harm Whats needed for coordination, well-differentiated roles for various actors, clear objectives & timeframes interventions experienced agency staff, with creativity and flexibility, funding and resources to work detailed contextual analysis, using existing information, learning from the past local consultation in needs assessment, willingness to work with non-traditional actors, low profile interventions safety and security of agency staff peacekeeping forces with protection mandate

consultations in needs assessment exercises, community self-targeting where possible, and a willingness to work with non-traditional actors. Safety and security of agency staff is of paramount concern, as is quick access to affected populations. Addressing root causes of the crisis is not a concern in such acute contexts, but there is an obligation to do no harm and this can pose difficult dilemmas for humanitarian agencies
37

which need to be approached from the standpoint of as good an understanding of the political economy and its dynamics as is possible. The presence of peacekeeping forces with a humanitarian protection mandate may be essential. Ituri, DRC as an example The group went on to consider a specific example, that of Ituri, DRC, drawing from the paper presented by Johan Pottier (see Box 5). Findings are summarized in Table 4. Salient features include widespread displacement and loss of life, the deliberate destruction of whole villages, and of livelihoods through crop destruction and loss of local markets/import markets controlled by militias. Land and cattle have been systematically stolen, and militias are adept at operating strategies of food aid predation. Thus farmers are unable to access fields and markets, local food availability is restricted by transport and market constraints, food aid is Table 4: Acute crises involving violent conflict the example of Ituri, DRC
Key features in Ituri Features: displacement loss of life (whole villages) deliberate destruction of livelihoods through crop destruction and loss of local markets/import markets controlled by militia theft of land and cattle strategies of food aid predation Constraints: farmers unable to access fields/markets problem of local food availability (related to transport/market constraints) food aid distributed and reclaimed by the army Possible potential: improved security in Bunia political future of warlords under discussion trade offs establish roles of different actors role of contextual analysis political/ecological clear objectives and clear time frames Safe havens What works Negotiation for return of lost assets Willingness to work with local actors Whats needed for interventions Analysis & information to work Low profile interventions Actors A growing list, including: UN agencies and peacekeeping forces missions, churches and religious institutions warlords child soldiers external players - e.g. Kagame, Museveni human rights agencies NGOs IDPs settled population government & alternative authorities military, merchants, politicians multinationals donors Importance of alliances these constantly shifting

38

distributed by the army only to be subsequently reclaimed. The deployment of UN peacekeepers holds out the prospect of improved security in Bunia, and the political future of warlords is under discussion amongst interested parties including the Ugandan army. Interveners need crucially to be aware of the many trade-offs involved and of the roles of different actors and of thorough contextual analysis, including political and ecological. Group 4: Lulls in conflict of uncertain duration (Table 5) This characterization applies to areas and times within most cases of protracted conflict, where fighting has been ongoing but is in recession for the time being. Parts of Sudan, much of Somalia, Eritrea during its long war of liberation, northern Uganda, parts of northern Sri Lanka and many other contexts would fit this model, which overlaps in part with the dappled conflict category considered by Group 2. In these areas of relative peace within what are described internationally as complex emergencies, many agencies are attempting to go beyond basic humanitarian relief to support what most people are doing anyway striving to rebuild and maintain their livelihoods. Yet this is not unproblematic, as a lull in actual fighting may coexist with a highly coercive, politically charged state of affairs, which may at any time return to all-out war. Key issues This group began with some observations on the humanitarian back to basics argument that until a complex emergency is over, interventions should not seek to broaden humanitarian objectives into developmental areas such as livelihoods support or rehabilitation, or into peace building, because neutrality and impartiality are likely to be sacrificed on account of the deeper involvement of local authorities that such interventions require. The argument suffers from the problem firstly that pure relief assistance also becomes politically embroiled in conflicts and affects their dynamics and it is difficult to make such sharp distinctions between relief, rehabilitation and development modes of intervention as these are objectives or outcomes which might flow simultaneously from the same activities (food aid distribution, for example). Secondly, it is often almost impossible to tell when a protracted complex emergency is over. Areas of peace within countries labelled as complex emergencies may persist for a long time, despite the lack of a national peace agreement (e.g. in northern and parts of southern Somalia) , or alternatively may re-erupt into fighting even after a peace agreement is signed, as in northern Uganda. The group also considered that a return to the pre-conflict situation is normally not an option and very often undesirable anyway, so there is a need to analyse the new situation and its implications for what options will work and how. What works will depend on the nature of the lull and of any likely resumed conflict, and the types of assets that could be affected. Finally, the group agreed that the subsistence farmer model underpinning many policies has limited validity. While crop production for own consumption remains a mainstay of food security for a large proportion of rural households in developing countries, there is a substantial dependence on other sources of food and cash to buy food even for those that do engage in agriculture, and much more so for agro-pastoral, pastoral and urban livelihoods. What works The group felt that situations of this kind call for interventions which either have a short-term pay-off for food security and livelihoods, or which establish a relatively protectable basis for longer term gains, or both. If interventions are to involve provision of assets, those less likely to attract predation or destruction are generally more appropriate. These might include cash, livestock if these can be protected, and in particular skills. Likewise, income generation activities with quick returns can help support coping strategies. Food aid can play a role as consumption support, or to allow rebuilding of herds and other assets, and food-for-skills and food-for-work programmes can provide longer term benefits.

39

Table 5: Protracted crises when there are lulls in conflict of uncertain duration
Humanitarian back to basics argument: true that pure relief wont influence conflict? Islands of peace in space/time can and should be exploited e.g. Sudan, Somalia Cant return to the pre-conflict situation so must analyse the new situation and its implications for what options will work and how Subsistence farmer model often/mostly invalid What works depends on nature of lull and any resumed conflict, and the types of assets affected Provision of assets (livestock, skills) which can be protected if conflict returns What works Building trust and relations with communities Income generation activities with quick returns to support coping strategies Food aid as consumption support, to allow rebuilding of herds and other assets. Can include food for skills and food for work. Integrated nutrition programmes to upgrade nutritional status Exploring longer term alternatives for marketing, livelihoods, productive infrastructure to take advantage of lulls, including agricultural and non-agricultural inputs Helping people return home when safe Build capacity of local agency staff Mediation and advocacy to assist resource sharing for improved food security Emphasising alternatives to conflict Livestock vaccination programmes What doesnt Fixed assets which are easy to attack (e.g. in N. Uganda) work Large scale & some mobile assets which are open to predation Complex interventions requiring international staff, likely to collapse if staff leave Interventions which can further war aims of belligerent authorities Continuity of agency presence helps to build trust with communities Whats needed for Understanding livelihood approach is essential for planning appropriate interventions interventions Understanding the conflict context helps to predict effects of interventions to work Information systems which help understanding of all aspects of the context Capacity (operational) to act on information in different timeframes and take risks Checklist of considerations (including ethical) to test options Coordination of approach between actors What actors Warring partners have the greatest capacity to disrupt interventions are involved Civil society actors have the best capability to lead/direct interventions in the absence of reliable state structures Affected communities The media Ethnic/religious/clan networks Local authorities External actors: donors and international agencies involved in food security Key issues

Integrated nutrition programmes may be needed for upgrading nutritional status, especially among children. In addition to provision of seeds, tools and other productive assets such as fishing nets, livestock vaccination programmes can provide valuable and relatively rapid benefits in support of agricultural and related livelihoods. At the same time there may be opportunities for exploring longer term alternatives for marketing, livelihoods and productive infrastructure to take advantage of lulls which can be expected to last. All such interventions can be linked with activities to build capacity of local agency staff. For the displaced, a priority will be to facilitate a return home when this is safe. This is likely to involve provision of food aid both in current locations, during return and for a period after

40

return until livelihoods can be re-established, plus livelihoods support in the form of inputs and productive assets, both agricultural and non-agricultural. The return home may, as in the case of Rwanda in the late 1990s, involve disputes over land and other resources as other users may have moved in during the period of displacement. Support for mediation and advocacy work may be necessary to help resolve such disputes. Similarly, the provision of assistance such as agricultural inputs or implements to people who may be occupying land vacated by those displaced by war requires careful consideration, as this may help entrench newcomers claims to land and intensify later disputes with returnees. Finally, the group suggested that interveners can seek and take advantage of opportunities for emphasising alternatives to conflict at various levels, including conflict over resources. What doesnt work Conversely, investing in fixed assets in areas subject to militia attacks and predation is likely to be unwise, not only because of a high probability that the investment will be lost but because such assets may actually attract destructive raids. Large scale and some mobile assets (e.g. livestock) may also be open to theft. Complex interventions requiring international staff are likely to collapse if staff leave when the security situation deteriorates. Interventions need to be kept simple, participatory, able to be sustained in the absence of international agency staff. Whether aimed at saving lives or livelihoods, some interventions can involve entering into agreements of various kinds which can further the war aims of belligerent authorities. Whats needed for interventions to work The working group stressed the importance of continuity of agency presence, even during periods of relative turbulence, to help build trust with communities though the balance between continuity of presence and protection of agency staff is not an easy one to maintain. The most productive approach to planning appropriate interventions was felt to be a livelihoods one, and understanding local livelihoods as well as the political economy and dynamics of the conflict is essential for appraising options and gauging their effect. Information systems can play a vital part in achieving this understanding if not restricted to the mere presentation of data. Finally, the operational capacity to act on information in different timeframes, and take risks in the process, is necessary. In comparing options for interventions, it helps to have a checklist of considerations or tests against which options are appraised. These need to include ethical considerations: is the proposed intervention likely to further the war interests of one or other party to the conflict in ways that will prove counterproductive to food security in the longer run? Like other groups, this one felt a need for the approach taken by different agencies to be coordinated between actors to avoid the development of programmatic inconsistencies. What actors are involved The most critical actors to take into account in choice of policy options are the warring parties themselves, or those who might resort to violence or otherwise pursue war aims within the area in which interventions are planned. These are likely to disrupt efforts to meet food security goals, or co-opt those efforts in ways that lead to unintended consequences for the beneficiaries or for other groups. In the absence of government agencies, or where the state as a belligerent party is likely to subvert or divert benefits of planned interventions, there may be civil society or communitybased partners which can play a valuable partnership or leadership role in programme implementation. Ethnic/religious/clan networks are often an important source of security and conflict management as well as channel for conflict itself, and yet tend to be ignored or bypassed by international agencies. Interventions which can work productively with such networks may stand a higher chance of success than those which do not.

41

Other key actors noted by the group include those targeted for interventions, donors, local authorities, other national and international agencies, and the media. On the last of these, the group observed that the way in which crisis situations are presented to the local and international media strongly influences popular perceptions, for better or worse. Public opinion in donor countries may indeed be a driving force for donors to act at all. Group 5: When longer lasting peace seems likely (Table 6) This group began with the suggestion that to identify what works it is necessary first to analyse the context, and further to specify for whom any given intervention works or does not work. Interventions can be expected to bring different degrees of benefit to different groups and may benefit some at the expense of others, especially where such groups are in mutual relations of conflict or subordination. Indeed this is one way in which interventions can be drawn into the dynamics of conflict as suggested in relation to Figure 2 above. The group considered food security as one of several linked aims contingent on sustaining peace, along with economic and livelihood security, reconstitution of institutions and policies, mitigation of conflict, protection of civilians and other objectives which are attainable only when peace lasts. Broader objectives may extend to policies to influence the course of conflict, strengthen the hand of government, other local authorities or external powers, or favour one group over another. Examples considered included Somalia, Somaliland, Angola, Sierra Leone, Liberia, Afghanistan, Iraq, Mozambique, Eritrea, Central America, East Timor. What works The group observed that with increasing likelihood of peace being sustained, a broad range of interventions with a longer-term perspective become appropriate and practically possible. Yet precisely because the success of these depends on sustainable peace, it is important to include policies and interventions which help to address issues that may be a basis for further conflict. The demobilization and reintegration of armed forces/militias and the return home of refugees and the displaced (major preoccupations of the newly established authorities in both Eritrea and Somaliland during the early 1990s, for example) are in themselves significant areas for potential international support, and can be linked to rehabilitation of productive assets by ex-fighters, but as noted for Group 4 they can also throw up difficult issues of land tenure which need to be addressed even during the transition to peace. The (re)establishment of livelihoods is a priority once fighting stops, and this can be assisted by a range of interventions involving, for example, support for credit and micro-credit schemes including loans to women, rehabilitation of infrastructure with an emphasis on local participation and inclusiveness, measures to secure better access to international and domestic markets, income generation schemes which are carefully targeted, sensitive to seasonal labour requirements and operate on sufficiently long time-frames, and support for diversification of livelihoods in order to enhance their sustainability. Donor support for post-conflict interventions of this kind is often disappointingly thin and short-lived, but both governments and donors tend in particular to ignore the needs of women and pastoralists and agro-pastoralists in rehabilitation programmes. The group drew attention to the need to rebuild herds after conflict, as in Eritrea where a combination of drought and war had decimated herds during the 1980s. Programmes to assist recovery in the livestock sector, which can include food aid as a means of reducing pastoralists need to sell animals to buy grain, can also help to defuse conflict between agriculturalists and pastoral communities. The status of livestock is important to include in post-war food needs assessments. Programmes to ensure the sustainability and improvement of both grassland and marine resources need to be considered, for example where war has meant a lack of effective regulation of access to coastal fisheries as in Somaliland.

42

Table 6: When longer lasting peace seems likely

Issues To work in what context? To work for whom? Food security as one of several linked aims contingent on sustaining peace (see diagram) A blend of actors involved Policies reflect broader objectives

Food security Econ security/ livelihoods

Conflict mitigation other aims Protection

Institutions Capacity building Policies

Examples Lessons

Somalia, Somaliland, Angola, Sierra Leone, Liberia, Afghanistan, Iraq Mozambique, Eritrea, Central America, East Timor

What is likely Credit, micro-credit, loans to women to work? Rehabilitation of productive assets involving youth (demobilization) Addressing land tenure, even during transition to peace Rehabilitation of infrastructure local participation, inclusiveness Restoration of livestock (food aid can help) and defusing conflict between agriculturalists and pastoral communities Food needs assessments which do not ignore livestock Sustainability and improvement of grassland/marine resources Negotiation for market access (international and domestic) Income generation targeting, accountability, time frames, seasonal calendar Support for livelihoods diversification to boost sustainability Bias towards emergency relief and food aid alone (unlinked to livelihoods) What does not work Sectoral bias e.g. crop vs. livestock systems Only government and administrative initiatives Top-down initiatives, without beneficiary involvement or local knowledge Engaging only elders/men Initiatives which are partisan or perceived as such Over-estimation of war-depleted government capacities Isolation of initiatives from peace processes Plans and processes reflecting external agendas, e.g. donor foreign policies Dealing with returnees in isolated settlements, apart from development initiatives Local authorities willing to pursue peace and accept international support Whats needed for Building of respect and trust with beneficiaries interventions Vigilance towards outside manipulators to work Coordination of donors and implementers Participative approach at all levels An adequate policy framework Reliable and shared information and analysis (baseline and variation, including shocks other than conflict, also in local languages) Mechanism for dispute resolution over resources/assets Supporting civil society elements (e.g. womens groups and INGOs) with an interest in sustainable peace Acceptance of links with protection What actors Armed actors sustainability of their reinsertion into communities are involved Private sector sustainable market entry points National authorities need rebuilding, an essential element of conflict resolution Local associations, civil society needs, priorities, sustainability Throughout consider gender, ethnicity, religion, traditions 43

What doesnt work Conversely, the group drew attention to the undesirability of a bias towards emergency relief and food aid alone unlinked to any strategy to rebuild livelihoods that often characterizes post-conflict interventions, along with a sectoral bias such as in favour of crop as against livestock systems. Similarly, initiatives which are top-down and involve only government and administrative structures rather than beneficiaries or local knowledge tend to fail. Engaging with community elders is important, but the participation of other groups including women is necessary too. Initiatives which are isolated from peace processes or which risk being seen as partisan are unlikely to succeed, as are those which reflect predominantly external agendas such as donor foreign policies. At the same time, there is often a tendency to overestimate the capacity of war-depleted government agencies to implement programmes without additional and long-term capacity-building assistance. Where returnees and demobilized fighters are concerned, the group suggested that attempts to house them in isolated settlements rather than integrate them into development initiatives are likely to fail, as was found in post-1991 Eritrea. Whats needed to make interventions work Consistent with its emphasis on the need for interventions to help sustain an often fragile peace, the group highlighted the prerequisite of local authorities which are willing to pursue peace and accept international support, as well as the building of respect and trust with beneficiaries and a vigilance towards outsiders who might seek to manipulate the situation to their own advantage in ways detrimental to longer term peace and security. Supporting civil society elements such as womens groups and NGOs with an interest in sustainable peace, and acceptance of links with humanitarian protection were also considered important. This group like others mentioned a need for better coordination among donors and implementers, a participative approach at all levels, and an adequate policy framework. Mechanisms for resolving disputes over resources and assets including land are essential in preventing a slide back into conflict. The group also drew attention to the need for reliable and shared information and analysis, both in terms of baseline knowledge of the context and up-to-date monitoring of events, including shocks other than conflict and including availability of information also in local languages. What main actors are involved The group made reference to a blend of involved actors, drawing particular attention to the sustainability of reinsertion of armed actors into communities, the need for sustainable market entry points for private sector actors, the need for a rebuilding of national authorities as an essential element of conflict resolution, and the attention to be paid to the priorities and sustainability of civil society actors including local associations. Throughout the programming and implementation of interventions, consideration to gender, ethnicity, religion and traditions is also essential. Conclusions from the exercise Participants agreed that this exercise highlighted the limitations inherent in attempts to generate blueprint policies for generic scenarios such as the five considered. Similarly, standardized responses based on agency guidebooks or experience from other cases cannot be unproblematically applied without detailed and critical consideration of their contextual appropriateness. In this sense the exercise bore out the need for the kind of flexible, light-footed analytic capacity called for in Doornbos presentation, and its effective linkage to action on the ground issues that were discussed on the final day of the workshop. The exercise did nevertheless highlight some key policy issues in planning short and longerterm food security interventions under different political and security conditions and which can therefore be regarded as elements to consider in building policy frameworks:

44

1. The significance of violent conflict. The capacity of violent conflict to transform capacities, assets and institutions on which food security depends means that the presence or absence of peace and respect for basic rights must be the overriding determinant of what is possible and appropriate in relation to food security. Since all interventions are likely to both influence and be influenced by the dynamics of actual or threatened violent conflict, their design must take into account not only the constraints that such conflict might impose on their implementation, but the manner of their likely incorporation into the conflict itself and the ethical and practical implications of this in terms both of possible complicity in human rights abuses and of achieving programme objectives. Donors must also accept that investing in interventions to promote food security in situations threatened by conflict is inherently risky. The argument underlying many contributions to this workshop is that not doing so involves an even greater risk to food security in the longer term. 2. The analysis of conflict. It follows from the above that intervention policies must be formulated in the light of not just the food security status and technical, economic and institutional context affecting target populations, but also of the military, political, economic, socio-cultural and historical dimensions of the violence that is ongoing or threatened and its regional and international aspects. This requires good background understanding overlaid with timely and spatially disaggregated conflict assessment, including assessment of actual or likely impacts of humanitarian and other interventions. Conflicts in complex emergencies rarely follow preordained trajectories from build-up to acute warfare to peace agreement and post-conflict recovery on the contrary they are unpredictable, may have a number of peaks which differ in their causes and effects, perpetrators and allegiances, and may drawn out over years or even decades. Undermining the resilience of food systems may be a deliberate strategy of subjugation, as Pottier showed for Ituri. 3. Identifying opportunities to boost resilience of food systems and help transform conflict. It is the combination of these kinds of analysis that is required to identify what kinds of opportunities can work in the sense of boosting the resilience of food systems while maximising positive and minimising negative impacts on conflict dynamics and human rights. The analysis needs to specify for whom alternative intervention options will work, and what will be the response of others for whom they might not work. Given the shifting, unpredictable nature of conflicts, it also needs to guide flexible, opportunistic policy making and programming. Multi-scenario planning methods referred to earlier may help build such flexibility into programming.8 4. Coordination between aid actors in developing policy frameworks. Even in crises which do not involve conflict, inter-agency coordination was recognized as a problem. Yet the fact that interventions to address vulnerability in complex emergencies also affect conflict dynamics which in turn determine what interventions are possible means that inter- as well as intra-sectoral coordination of assistance becomes even more essential than in non-conflict situations. Within the UN family of agencies there are signs that under the leadership of the UNDP Resident/Humanitarian Coordinator supported by OCHA, progress has been made in establishing common policy frameworks at country level which recognize and build upon the mandates and core commitments of individual agencies. Such frameworks may include international NGOs and even some bilateral donors. But bilateral actors also have foreign policy agendas in relation to complex emergencies e.g. pursuance of the War on Terror, limiting the influence of particular regimes, the containment of potential refugees, or securing strategic access to key
8

Multi-scenario planning or simply scenario planning has found increasing application in both public and corporate spheres over the last 15 years as an approach to planning in uncertainty. Pioneered in the 1980s by Royal Dutch Shell, it was famously applied in the Mont Fleur project in Cape Town in the early 1990s. The project involved an extended series of brainstorming sessions aimed at establishing a set of alternative scenarios or development pathways for the future of South Africa on its transition to democracy.

45

mineral resources which may be inconsistent with humanitarian principles and which they may be unwilling to submit to inter-agency coordination, and yet the effects of these agendas must still be factored into common policy frameworks. 5. The content of interventions for food security. The overarching food security policy aim is protect, sustain and promote food access in the face of critical threats. Interventions which might be most appropriate in our different conflict scenarios include the following:
Where widespread political violence is not a factor broadening access to agricultural

services, inputs and internal and external markets can be combined with promotion of non-agricultural livelihoods and the design of social protection measures along with carefully selected targeting mechanisms for different groups in most critical need.
With the threat of conflict resulting from a deteriorating political situation these same

policies need to be supplemented with special efforts to defuse political tensions, ensure continued functioning of markets and infrastructure, protect assets against liquidation especially amongst the most marginalized and boost the resilience of agricultural systems. Conflict early warning systems offer the prospect of more timely and more complete analysis of deepening tensions and the emergence and dynamics of violence.
Acute humanitarian crises resulting from an outbreak of widespread political violence

indicate a need for humanitarian protection including safe havens, negotiation with belligerent authorities for humanitarian access possibly but not necessarily including food aid and nutrition support, plus other lower profile interventions which do not attract predation and agricultural inputs such as quick maturing seeds for food and cash crops and other productive assets. Agency staff security is likely to be a major preoccupation, but a priority is also to avoid compromising humanitarian access and unintentional complicity in abuses.
In areas and times where there is a lull in violence but it is not clear how long this might

last, there may be opportunities for short-term support for food security and livelihoods and for interventions which provide relatively protectable assets as a basis for longer term improvements in resilience, such as cash, possibly livestock and livestock vaccination, and skills. Income generation activities and support for local institutions may also be appropriate. Again, care is needed to ensure that interventions do not strengthen the hand of belligerent forces, but help where possible to protect the vulnerable and generate incentives for peace.
The prospect of a more lasting peace opens the way for measures to assist displaced

communities and demobilized fighters to return home, rebuild diverse livelihoods and agricultural and food systems when they get there, and survive in the meantime. Credit schemes, the reconstruction of marketing infrastructure and the reconstitution of services including health and education are likely to be needed, along with measures to entrench peace and promote reconciliation. Special measures to support marginalized groups including women, female ex-fighters and pastoral and agropastoral systems may be required. Land tenure is a key issue in building food security, and a potential source of further conflict. Care is needed to avoid legitimising contentious land claims. Though post-conflict government implementation capacity may be weak, high profile interventions implemented through government structures can help to legitimize a newly established government.
Participants felt that the often short-lived donor interest in post-conflict rehabilitation is

regrettable: important opportunities for recovery of food security and livelihoods and avoidance of further conflict are missed. In particular, donors seem to show little interest in land tenure issues, even though this is a vital aspect of post-war recovery. 6. Other cross-cutting issues applicable to all of these scenarios are the following:
Food aid potentially has an important role to play in all of these situations. Where

coping strategies have failed and people are in a situation of acute food crisis there is no substitute for humanitarian relief which often needs to include food aid in some form
46

 as free distribution, food for work, supplementary feeding, school feeding etc. But food aid brings with it risks associated with its capacity to alter markets, prices and incentives and undermine livelihoods, and hence needs careful targeting and exit strategies. Interventions which involve the participation of beneficiary communities stand a better chance of being valued and sustained, though participation in acute situations may be wishful thinking. Rural-urban linkages are vital for livelihoods and food security, as is food access for urban populations.
HIV/AIDS issues run through all these scenarios and affect the suitability of

interventions. Violence, displacement, destitution and militia activity are all associated with increased exposure to HIV. Interventions need to take into account the requirements of agricultural and livelihood systems adapted to reduced labour inputs on account of AIDS mortality and morbidity.
Households do not exist in isolation from social and cultural networks including those

based on ethnicity, clan, gender and religion. While sometimes a source of conflict these networks are also vital sources of resilience and need to be supported as such.
Accountability and learning support is an important feature of policy frameworks in

crisis situations. Despite the uniqueness of complex emergency contexts, they do offer lessons which are relevant to if not applicable in new situations. There is a need to ensure that accountability is maintained and important lessons learned. The Learning Support Office piloted by ALNAP in Malawi could perhaps provide a useful model if adapted for complex emergencies.
Early warning and information systems which are strongly linked both to background

research and to policy formulation were regarded as important in all scenarios. Strong information-knowledge-action links are essential for ensuring that responses remain attuned to changing contexts, but the necessary research, information and communication systems for achieving participative knowledge generation are very difficult to set up during acute crises and therefore need to have been established in advance. These issues were considered in more detail on the last day of the workshop.

47

Theme 3: Implications and next steps

Introduction
Theme 3 aimed to consider implications of the issues discussed under Themes 1 and 2 for research and information systems and their role in policy formulation and programming. A panel of participants involved in food security information systems and related research made plenary presentations, followed by a presentation on proposals for a FAO information portal on food security and protracted crisis. Group work then focussed on distilling conclusions from the workshop. Participants completed a workshop evaluation form at the close of the final day.

Session 4: food security information systems in and out of crisis

The first session of the day considered implications for food security information systems and their role in policy formulation and programming. A plenary panel made four presentations on different aspects of food security information systems9. Food security policy in complex emergencies: information, analysis and programming in the Horn of Africa Samuel Zziwa, Chief of Agricultural Development and Food Security at the IGAD (Intergovernmental Authority on Development) Secretariat in Djibouti, gave a presentation of early warning work in the IGAD region. He outlined three types of food insecurity prevalent in the region:

chronic food insecurity, where adverse ecological and economic conditions prevent
communities sustainably producing or purchasing adequate food;

creeping food insecurity: progressive deterioration in ecological and/or socio-economic

conditions resulting in rising incidence of undernourishment and malnutrition.

transitory food insecurity caused by temporal shocks, mostly droughts and floods and less
commonly migratory pests especially locusts, periodically assuming disaster proportions and often involving displacement of populations internally and cross-border. At least 20 million people out of the regions total of 160 million are food insecure and over 5 million normally require food assistance. Areas facing severe food insecurity have also tended to become embroiled in protracted conflicts. The Horn of Africa is not only one of the most food insecure regions of the world, it is also fraught with conflicts of all sorts: tribal, insurgency, cross-border, cattle rustling, etc.. This has prompted IGAD to establish a formal mechanism for conflict early warning and response as a complement to the existing early warning and food information systems. Zziwa expressed regret that poverty reduction strategy papers (PRSPs) being prepared by IGAD Member States are still weak on food security. There is need for a holistic approach to
9

Time constraints prevented presentation of the paper by Gnter Hemrich on Matching Food Security Analysis to Context: The Experience of The Somalia Food Security Assessment Unit. The Unit, run by FAO with EC and USAID funding, is the main source of food security information for the SACB and other users. Its 22 field-based monitors report to a core unit in Nairobi, which uses a household economy approach to analyse and disseminate findings on key indicators for crops, prices, rainfall, external trade, nutrition, food aid and livelihood, humanitarian and food security trends. Hemrich analyses some key issues that the FSAU is grappling with, including the extent to which the Unit should engage in conflict analysis, fluctuation between short and longer term information needs, barriers to gender analysis, the profile and duties of monitors, and the Units remit vis--vis the SACB. (See Volume 2)

48

addressing food insecurity in the region. The strategies should cover the entire food chain from production, marketing to consumption. IGAD Member States have established and improved early warning systems which focus on transitory food insecurity, but they are weaker at detecting and analysing creeping food security. Systems provide seasonal forecasts of crop and pasture conditions, as the season progresses, using data from national meteorological and remote sensing services assisted by the Nairobi-based Drought Monitoring Centre. Against background data supplied by the IGAD Crop Production Systems Zones mini-GIS (CVIEW), these data make localized crop production forecasts possible. These production indicators need to be complemented with monitoring of indirect, and where possible direct, indicators of marketing and consumption. From the many indicators that can be monitored, a few most pertinent ones that can be sustainably followed within budgetary and cost-effectiveness limits need to be selected. From experience with regional food security monitoring, the IGAD Early Warning and Food Information System (EWFIS) proposed the following minimum data set for continuous monitoring of current vulnerability indicators at the local level:
meteorological data (rainfall, soil moisture, Water Satisfaction Index) remote sensing images (vegetation cover, rainfall estimates) agricultural statistics/forecasts (planted area, yields, expected production) wholesale commodity prices (food crops, livestock, non-food inflation) clinical child nutritional anthropometry (age under 5 years, weight, height).

With FAO support IGAD has developed an internet-based marketing information system, whereby national marketing agencies supply weekly price data for main food staples in major towns. Arrangements are underway to strengthen the livestock component, enabling monitoring of terms of trade between food staples and livestock to determine the status of the purchasing power of pastoral households in remote areas. There is a need for support to refine and systematize collection of clinical anthropometric data for use in generating locallevel trends of nutritional and health status among under-5s for early warning purposes. National early warning systems are backed up by global ones such as FAOs GIEWS (Global Information and Early Warning System), WFPs VAM (Vulnerability Analysis and Mapping), and USAIDs FEWS (Famine Early Warning System). Most critical are the FAO/WFP-led inter-agency crop and food needs assessment missions carried out at the end of the main growing season in most Horn countries, providing the basis for appeals for food and other assistance where necessary. Despite advances in information technology which have revolutionized early warning and food information work, violent conflict imposes severe constraints on collection of food security data. In most such situations, however, some baseline information already exists which can be used to guide interventions. Such information relates to the resource situation, land tenure, potential and use, accessibility and infrastructure, social systems and dynamics, agricultural and livelihood systems, nutrition and health status etc. This information can help build into relief programmes the means to facilitate recovery through agricultural rehabilitation, packages to assist the return of the displaced, creation of non-agricultural employment through appropriate income-generating activities including work-based programmes to rebuild infrastructure, and the exploitation of new opportunities that the post-war situation offers. It can also provide a basis for establishing conflict early warning systems that can warn of nascent hostilities and prompt preventive action. In particular, even while violence rages, a systematic effort to collect information from displaced persons (who know their area best) can help inform policies and programmes to address recovery and sustainable development issues when the conflict comes to an end.

49

Food and conflict early warning systems in complex emergencies Dan Maxwell, Regional Food Security/Livelihoods Advisor, CARE International in Nairobi reflected on the challenges that complex emergencies pose to early warning systems, and debates about whether the focus should be on risk factors or food security outcomes. He felt they should tackle both. Needs assessment should consider all possible outcomes. Links between baseline vulnerability assessments, early warning systems and action are all vital. A key challenge to early warning systems is how to include political analysis. Recognizing conflict as an affecting factor is one thing, but predicting events is another matter this calls for conflict early warning systems. Yet these are politically very sensitive. IGAD has its Conflict Early Warning (CEWARN) project, but so far this considers only pastoral areas along inter-state boundaries. Conflict early warning is a problematic area. How should it embrace neutrality? Should it use an approach based on indicators? Qualitative assessment would probably be more effective. Traditional early warning systems look for problems (e.g. threats to food security), but conflict early warning systems are also about identifying opportunities to mitigate or prevent conflict. In this regard Maxwell referred to the tendency for OCHA/UNDP contingency planning exercises to adopt a rather doom and gloom perspective on conflicts which is not always warranted. In relation to the conflict in Burundi, for example, it emerged at the recent ICG meeting in Bujumbura that there was space for mediation and action to mitigate conflict as both main parties had overplayed their hand. One approach to measuring resilience is the Coping Strategies Index (CSI) developed by CARE and WFP in East Africa. This uses a scoring system to assess answers to the question What do you do when you dont have enough food, and dont have money to buy food? A range of alternative options, ranking in order of severity from rely on less preferred and less expensive foods to skip entire days without eating, is provided, and scoring based on the frequency each strategy is employed and its relative severity. Monitoring whether the resulting index rises or declines gives a rapid, real-time indication of whether household food security is deteriorating or improving. This simple and well tested tool can be used as part of either early warning or assessment, and in conjunction with food aid end-use data can give a rapid indication of food aid impacts. Nutrition and food security information systems in crisis-prone countries Nutrition information provides one of the few means of evaluating peoples overall well-being, and the impact on them of a crisis and of interventions to address it. It prompts an analysis that is intersectoral and goes beyond food availability, and supports an evaluation of concepts of normal food security. Measuring the nutritional status of young children during a crisis is not only a basis for specific measures for their benefit, but is also a tool in evaluating the welfare of the entire population. Further, within a multisectoral data collection and analysis system a nutrition focus can yield an understanding of the reasons for crisis impacts on a particular population, which is more useful in the longer term. Noreen Prendiville, Nutritionist in the Somalia FSAU, outlined findings of a comparative study of nutrition information systems in Somalia, Sierra Leone, Eritrea, Ethiopia, Afghanistan and Kenya, conducted by the FSAU in 2002/03. The aim was to gauge their effectiveness, in the context of the food security and other related information systems in each country, in meeting decision makers information needs during periods of crisis. Major findings were as follows. Outside times of crisis

Government-run systems are generally weak, collect information only on chronic

malnutrition, and do not transmit information regularly or manage data effectively. Data is often summarized routinely but then hardly used. In general, data remains within the

50

health sector and little or no use is made of it by those producing the information at health facility level.

The externally-funded Multi-Indicator Cluster Surveys and Demographic and Household

Study provide very useful national level information allowing comparisons in nutrition status between major administrative areas. But these surveys, undertaken every 2-3 years, will not detect crisis-related change over a short time period, or identify specific vulnerable populations. Moreover, the most crisis-prone areas of a country are often also the most inaccessible, so that information from these areas is generally of poorer quality.

Intersectoral sharing is rare: typically nutrition information (mostly anthropometric data)

remains in the Ministry of Health while food security information (mostly food availability focussed) remains in Agriculture and related ministries. Separately, these two types of information cannot provide an adequate understanding of dynamic impacts of a crisis on the welfare of a population.

Governments use both food security and nutrition information as part of their core national
level monitoring through central statistics offices. International organizations also use this information for monitoring country progress on a global level. Area-based systems are supported by individual NGOs (e.g. SC-UK in Ethiopia) which may use state-of-the-art methodologies to support projects but do not aim for sustainability or countrywide coverage. However information tends not to be used outside times of crisis for analysis that could improve crisis preparedness and response. During periods of crisis

These inadequacies mean that at the onset of crises ad hoc information systems are set
up in parallel with existing ones, and little time is available to forge links between the two. When the crisis recedes the emergency system tends not to be maintained. Apart from in Somalia, lack of baseline information means that the real impact of the crisis cannot be accurately evaluated.

The focus is on short-term needs of humanitarian operations, in particular identifying

pockets of acute vulnerability (with relative neglect of other areas) and stepping up nutrition surveys using indicators of acute malnutrition and wasting. Food security information systems are established or enhanced, using more sophisticated methodologies such as vulnerability assessment and mapping or household economy analysis. There is a strong reliance on external funding and expertise, and information leaves the country during or immediately after the crisis.

Sharing of information occurs, but mostly among international agencies involved in crisis
response. Towards more effective nutrition & food security information systems in crisis-prone countries The significant external support for information systems during crises provides an opportunity to invest in sustainable, countrywide systems that serve both short- and long-term information needs, integrate food security and nutrition aspects and benefit from longer-term donor funding commitment These should be developed within country-based structures, use transparent, easily understood methodologies tailored to country needs and less reliant on quantitative indicators, and involve much improved national staff training. Such systems will help reduce vulnerability to future crises by promoting an understanding of dynamic shifts in livelihoods occurring in many crisis-prone countries. Information and knowledge in complex emergencies Sue Lautze presented a paper which made four main points about the different status of information and knowledge in the humanitarian system.

51

1. We mustnt automatically assume that the system is so broken that there is adequate will to fix it. It is critical to analyze for whom the current system is functional: for most humanitarian agencies and their staff, and for donors for whom the political imperative is to be seen to be generous to humanity in extremis, the system works well. 2. Yet the system is not serving the needs of disaster-affected populations, and this is due in part to the humanitarian communitys rather structural aversion to knowledge and near addiction to data, information and quick, technical fixes to deeply complex problems. 3. The most pressing area for improvement, in the realm of knowledge, is to understand the relationships between violence and vulnerability on many levels but most particularly how violence affects societies. We must invest in building analytical frameworks, in conducting the research, and in analyzing data and information in order to gain a much better appreciation of how violence threatens and changes policies, institutions and processes and, in turn, how the formal and informal policies, institutions and processes of societies demonstrate resilience in the face of such violence. The humanitarian community nibbles at the edges of this question by studying coping strategies and engaging in capacity building, but this is a far cry from countering the myriad of assaults on states, communities, families, policies and traditions inherent in todays complex emergencies. 4. Finally, we need to rethink what it means to be humanitarian. The very word is based in humanity, in the individual. To serve the individual requires humanitarians to engage with the guys with the guns engage in the local, national and global processes that are shaping the vulnerability environment in which far too much of humanity exists. Discussion Discussion focused on three areas: incentives for information versus knowledge generation, international processes in response to emergencies and the role of food security early warning systems in this. The importance of making early warning systems as neutral as possible was noted; however donors often trust only their own sources. The linking of food security and conflict early warning systems holds out the prospect of better political analysis in complex emergencies something that is often called for. Yet there is little concept of how the two kinds of system might be linked, and conflict early warning systems need more work to overcome significant methodological constraints. On indicators, it was observed that information on so-called early warning indicators often comes in at a late stage, and is often not adequately related to background information and knowledge. Also, the shortlist of key indicators that need to be monitored can differ significantly from case to case: in Somaliland, for example, an understanding of the food security situation requires monitoring of access to marine resources, trade in khat and charcoal, informal cigarette imports and governance as well as those in the minimum list. A strong on-going engagement based on local networks was considered important to improve monitoring of food security trends, albeit with a wary eye to the role of deliberate disinformation. Concerning incentives and international processes, it was noted that while there is little cause for optimism on a shift in emphasis from information to knowledge, the impetus for such a shift may come from donors. A better understanding of different donor, international agency and NGO perspectives is required to understand information needs. FAO information portal on food security and protracted crisis Jacky Sutton, Information & Communications Consultant in FAOs Food Security and Agricultural Projects Analysis Service, presented a pilot version of a new International Portal on Food Security and Crisis in Countries Subject to Complex Emergencies, a recent FAO initiative that was expected to be launched before the end of the year. The site would have three main goals:

to provide relevant information to policy makers and planners as well as field personnel

52

working in complex emergency environments;

to provide a forum for food security analysts, policy-makers and field-level actors through
email listservs and online discussion groups;

to highlight institutional initiatives to develop food security analysis linking emergency

relief with long-term and normative development. A brochure on the proposed portal was circulated. Soon after the workshop the portal went online at http://www.fao.org/crisisandhunger.

Session 5: Conclusions on principles and agendas for action

Working groups then considered conclusions of the preceding workshop sessions for principles and agendas for further action, leading to a final plenary report-back and discussion before the workshop was brought to a conclusion. Groups were based on four different perspectives or constituencies:
Group 1: Website and networking Group 2: Research agendas Group 3: Policy principles Group 4: Food security field programming

Group 1: Website and networking Usefulness of FAO portal/website The group considered the portal initiative useful. It would enable FAO to provide information, resources and a discussion forum with a specific focus on food security and conflict and links between the two. It would differ from existing information sources such as ReliefWeb in providing a comprehensive database of background documentation relevant to researchers and policymakers concerned with impacts of conflict on the food security situation and issues related to the transition from relief to longer term food security. It would have the potential to act as an honest broker of information in situations subject to deliberate strategies of disinformation. The ICRC delegate in the group considered that ICRC would be interested in the food security analysis provided by the portal as a complement to conflict analysis and to provide a baseline understanding for its field staff. The group underlined the importance of an institutional commitment from FAO to update and maintain the site. There would be advantage in operating moderated online discussion groups, but a need to be wary of too much dependency on one appointed moderator. Comparison was made with the experience of two information systems operated by IGAD:
the Regional Integrated Information System (http://igadriis.uonbi.ac.ke) which is a

clearing house and search tool in its pilot stage, with a custodian institution (University of Nairobi) sending information to a gateway server or three regional nodes;
the Market Information System (http://www.igadmis.net), which provides information on

market prices, updated weekly: the key challenge with this is to persuade producers that it is to their benefit to disseminate information via the site. The group recommended operating the site for a trial period, perhaps 6-12 months. Networking The group considered the prospects for establishing one or more international networks to support improvement of information systems, decision-making, inter-organizational discussion and field-headquarters dialogue. This is linked to the question of a possible series of workshops, perhaps on a regular basis, to follow the present one. Participants were interested to maintain a dialogue on the issues raised at this workshop, and discussed questions as to networking objectives and whether individually or institutionally based.

53

The group made three broad recommendations concerning action for networking involving institutional commitment and collaboration in areas of specific interest in an information context: 1. Establishment of a peer review group to monitor standards (content, relevance, quality) of information and collaboration. 2. Creation of a set of guidelines/indicators to identify/measure/analyse social stress and ways to intervene in policy arenas through advocacy. These would bring together disparate indicators (e.g. press freedom, livestock sales, remittances) and aim to promote an inter-agency voice which experience suggests is more effective at resource mobilization than agencies acting in isolation. 3. Ad hoc collaboration on specific, technical issues, e.g. the design of a country-wide assessment and aid coordination system for food security (learning from the experience of FSAU and SACB in Somalia); an agreed methodology/framework for technical reporting that can be easily integrated into a countrys infrastructure (lessons to be learned from Ethiopia here). Group 2: Research agendas The group identified key focus areas for research relating to interveners and contexts: Interveners
Further analysis of donors and their behaviour in relation to food security in complex

emergencies
A review of humanitarian mandates and how they relate to practice, and examination of

where mandates have become subsumed by other agendas such as peace building
Improving the research capacity of field agencies, including capacity to build on local

knowledge and understanding

A study of approaches to coordination in emergencies Learning from success conflict diversion and containment Exploring factors determining institutional knowledge and memory.

Contexts
Improving our understanding of contexts and what lessons can be learned from them,

including issues of developing methodologies for contextual analysis and refining concepts
Tapping into the knowledge of researchers for specific areas Assessing and improving indicators for access and vulnerability Exploring the dynamics and causal relationships between food insecurity on one hand

and conflict and conflict resolution mechanisms on the other

Analysis of relationships between the illegitimate state and war economies Understanding the dynamics of livelihood systems under stress, including how coping

mechanisms are able to adapt to longer term stresses such as HIV/AIDS

Use research to draw attention to the disconnect between the problem and the

response, and understand failures to make effective use of information available from many sources.

54

Group 3: Policy principles Insights The group identified a range of insights for policy formulation that had emerged from the exercise of attempting to identify what works and what doesnt work in different conflict contexts.

Policy is not something that can be developed in a generic sense, since it is:
institution-specific, but linked to institutional networks context- and process-specific: it must adapt to contexts which change both predictably

(proactive planning) and turbulently (reactive flexibility), and seize opportunities for innovation
an iterative process: policy must be revisited often and take into consideration possible

impacts of and on long-term processes

Policy must be based on knowledge-generation capacity built into information systems

and tap into knowledge bases, including local ones: there is still a large gap between high quantity of information, appropriate level of knowledge and decision making processes.

There is a need to identify constraints and opportunities beyond the emergency (including
islands of peace) and dynamics for change

Policy must be transparent and disseminated widely including in local languages Policy must recognize differences and similarities between local and institutional priorities
including gendered ones

Policy needs to reflect dynamics between violence and food security When failures of political systems, institutions and livelihoods have led to crisis,
addressing those failures should be the prime objective rather than resilience per se. Concepts and principles In the light of these discussions the group made a range of observations on concepts and principles for policy frameworks:

There is a need for further clarity on what food security is in complex emergencies, and
how it relates to other objectives such as protection and livelihoods

The concept of policy frameworks needs further definition and deepening: a policy
framework is a set of principles, approaches, goals and guiding questions for designing policies for courses of action not a one size fits all prescription

Constructing a policy framework means:

clarifying who the policy framework is for and what is its purpose or function identifying what ethical and operational principles are applicable: these must be

relevant, sound and responsive

establishing and addressing guiding questions in relation to, for example:

the impact of violence on food security the dynamics of markets what options there are for assistance and asset building what needs for protection exist the variable strengths of local institutions time frames and the sequencing of interventions the balance between pro-active planning and reactive flexibility involving appropriate actors in designing and responding to such guiding questions.

Policy frameworks should prioritize what is a priority to local people this means:
55

 monitoring changes in local priorities looking at what hinders livestock survival (e.g. cultivation and use of khat or charcoal

production for urban use)

supporting grasslands and marine resources management and protection identifying local peaceful pockets threatened by administrations and major clan powers,

and then think of how to help manage peace

consulting women as leaders everywhere and anywhere communicating in local languages integrating marginal communities and seeing policy formulation as a challenge to

power networks not limiting policy to emergency areas

time matters creating functional, credible knowledge networks in advance of crisis deciding on the when, how and with whom of participation is still a complex

requirement in crisis

Policy framework formulation needs to be followed up with training, advocacy and

dissemination. Donts The group also warned against:

relying on blueprints based on general concepts not rooted in practicalities assuming that elders and administrators are fully representative of the most vulnerable.
Group 4: Food security field programming Principles for food security programming The group drew attention to the following guiding principles for field-level programming that emerged from the Theme 2 group exercise:

How to achieve beneficiary involvement: some tools exist to obtain relevant information on
priorities of different local groups, but getting operational actors to use them raises issues of capacity building, use of field officers, resources, time etc.

Coordination of information gathering, analysis and action raises further issues of what,
who and how there is a need to understand underlying causes, effects and dynamics at local, regional and international levels.

Protecting people and their assets is a key objective or principle of programming Do no harm within the bounds of humanitarian imperatives, trying not to make matters
worse

Seeking to empower national and local institutions to be accountable to humanitarian

principles and to formulate and implement policies that reduce vulnerability

Optimising

impact through clear, prioritized objectives, taking into account implementability relative to the scale of needs but investing in building capacity to expand scope and outreach

Integrate monitoring and impact assessment into programmes Flexible funding to allow rapid responses matched to context Programming with a longer term vision.

56

Beneficiary and institutional involvement The group gave further thought to involvement of beneficiaries and institutions in programming:

Why?
To get better information, deliver appropriate responses and build capacity Promotes ownership, commitment to actions, contributions, sustainability Promotes legitimacy of local institutions and formulation (internalization) of more

appropriate policies

What?
Budgetary support for institutions, including unearmarked funds disbursed against

agreed principles
Institutions include CSOs, NGOs, local authorities and national governments.

How?
An inventory of whos who, doing what and where Dialogue/facilitation with representatives of existing networks and groups of the

vulnerable
Capacity building of local actors and training of managers and field staff Dissemination of information on objectives and processes An ongoing process involving monitoring and field based research

Examples The group considered the following positive examples:

SACB Somalia: resource people carrying out participatory activities Afghanistan: Ministry of Rural Rehabilitation and Development and Afghan Assistance
Coordination Authority

The Somali Red Crescent Society The ALNAP Learning Support Office in Malawi The revised Sphere standards for humanitarian assistance which now include food
security aspects

The Global Study on Participation being conducted by Groupe URD and ALNAP
(www.globalstudyparticipation.org)

Groupe URDs Quality Project to improve the quality of interventions in emergencies

(www.qualityproject.org) Closing discussion The policy-conflict matrix emerging from the Theme 2 group exercise (Error! Reference source not found.) was reflected upon. While not usable as a policy blueprint for different generic situations, it did offer a useful framework for discussing food security policy and programming issues from a conflict perspective, suggesting those issues that are likely to come to the fore or options that might need special consideration with differing degrees of proximity to violent conflict. Those discussions highlighted how violent conflict tends not only to impact severely on food security and complicate assessment work and the delivery of interventions, but also to transform the entire arena political, institutional, socioeconomic, ethical as well as technical for making decisions about responses and gauging likely outcomes. The cases considered demonstrated further that there is no predictable trajectory taking complex emergencies through successive stages from threatened conflict through an
57

acute phase and finally into a post-conflict status rather contexts can change unpredictably and rapidly while remaining in crisis seemingly indefinitely. It was noted that policy processes in emergencies are often less than rational, having their own political economy not necessarily based on objective information. It was recalled that in armed conflicts, the principle of giving primacy to making peace is sometimes set aside, not just because of a lack of political will to address it, but also because there is no means of setting priorities between reducing conflict and other humanitarian interventions. Donors reluctance to commit to longer-term funding for rehabilitation was also highlighted. These difficulties in part reflect a lack of mechanisms for setting priorities in protracted crises which address peoples own priority concerns in both short and longer term. Implications for policy frameworks for food security in complex emergencies are that they need to:

extend to all aspects of food security, not just food aid and agriculture, and link short-term
interventions to longer-term planning frameworks;

promote capacities for policy-making that is responsive, flexible, opportunistic and

reiterative, and closely linked to systems for information and knowledge generation while inflexible policies determined in advance could well have a negative impact;

include stronger and more transparent ethical and operational principles which can take
account of how interventions are likely to be incorporated into and influence the dynamics of conflict. Questions were raised as to the extent to which donor conditionality and earmarking of funds were driving interventions. Donors have espoused a principle of withholding development aid but funding humanitarian assistance in countries in which authorities are involved in conflicts, yet the amount and type of humanitarian assistance provided tend to be conditioned by donor foreign policy objectives, and the limits of what constitutes humanitarian assistance have become increasingly blurred. Capacity-building has become an important aim for international agencies, yet in conflict situations the question of whose capacity is to be built is often a difficult one. The role of the ICRC and International Humanitarian Law (IHL) was raised in relation to such dilemmas. While the ICRCs mandate and legal status mean it is neither an intergovernmental organization nor an NGO, it must establish agreements (subject to IHL) with states or authorities for the independent provision of humanitarian assistance. This means there is often a need to engage with the guys with the guns, something which can only be practically achieved from a position of neutrality and impartiality. This is however not inconsistent with ICRCs activities in supporting livelihoods and building capacities in, for example, the health sector. At the same time it must be accepted that humanitarian agencies often respond as much to the CNN factor as to early warning system information. Finally, reference was made to the importance of the right to food, during emergencies as well as at other times. In the absence of a functioning state, the question of who should guarantee this right presents particular difficulties. In summing up the workshop, Hartwig de Haen, FAO Assistant Director-General, Economic and Social Department, congratulated participants on bringing out the very complex nature of food security issues in crises and the challenges they present, and the absence of simple solutions or quick fixes. Food insecurity was considered both an outcome and a cause of conflict and highly political. Thus vested interests would continue to be strong, but better understanding and knowledge of links could be one of the keys to improved food security. In his closing remarks, Prabhu Pingali thanked the participants and organizers, emphasized the importance of food security analysis in relation to complex emergencies and highlighted this as one of the key ESA research areas for the coming years.

58

Annex A. Policy conflict matrix

When there is the threat of violent conflict involvement of local actors markets basic support to agricultural production coordination disarmament, demobilization, reintegration (DDR) conflict impact analysis establishing basic services - theatre of normality actions to reduce destitution/decapitalization participatory work targeting (versus blanket coverage) reinforcing capacities of institutions create employment - even temporary involvement of diaspora food aid improved and enforced access to natural resources early warning it depends. market interventions to lower food prices? dialogue/political processes - but not by FS agencies necessarily working through local leaders/merchants for delivery advocacy for political processes & incentives for warlords to allow access quick maturing vegetable seeds - to promote via veg. markets build on existing programmes & structures different distribution & targeting techniques use of cash to induce markets negotiations for return of stolen cattle & links to animal health safe havens wet feeding (cooked food) food aid (even though its reclaimed) therapeutic & supplementary feeding protectable assets agricultural inputs mediation for resource sharing integrated nutrition medium to long-term views income generation activities with quick returns building capacities training/planning food aid as food for skills/work; food for consumption support; food for rebuilding assets helping people return home (when safe) exploring alternatives to conflicts Acute crisis involving violent conflict Lulls in conflict of uncertain duration When longer lasting peace seems likely rehabilitation of infrastructure (breaking remoteness) address land tenure during transition already adapt food aid policy to transition (restoring livelihoods) livelihoods diversification (livestock, marine, nonagro) credit - micro-credit market access negotiation (international & domestic) income generation initiatives (livelihoods, diversification) productive asset rehabilitation involving youth (demobilization)

Multidimensional crisis without conflict

What works?

measures to increase access to agric. inputs vouchers, fairs less labour intensive/low input technologies access to services private sector enabling high asset build up/safety nets for orphans & elderly measures to increase effective demand: cash/food for work; asset creation; flexi food voucher school feeding grain reserves - national and regional; grain futures innovative credit reduce physical & fiscal barriers to trade

Multidimensional crisis without conflict involvement of diaspora knee-jerk rehabilitation reliance on markets only illegitimate, badly managed coordination mechanism DDR targeting (versus exclusion/inclusion) food aid it depends . unprotectable assets (large scale/ mobile) expatriate-dependent interventions interventions at risk of manipulation by belligerent authorities

When there is the threat of violent conflict

Acute crisis involving violent conflict

Lulls in conflict of uncertain duration

When longer lasting peace seems likely insufficient participation (imposed from outside, insufficient consultation) external agenda (e.g. foreign policy) over-estimating wardepleted govt capacities bias/restrictions in interlocutor selection addressing resettlement/reinsertion in isolation from development efforts emergency relief bias partisan initiatives (or perceived as such) sectoral bias (e.g. crop vs. livestock systems) isolation of proposals from peace initiatives food aid alone

What doesnt work?

free distribution of inputs business as usual agricultural extension self sufficiency focus relying only on the market over-complex projects

Multidimensional crisis without conflict more local purchases appropriate public information improved early warning develop legal instruments to improve access to resources improve the FS information system better understanding and support of social safety nets good institutional analysis appropriate feasibility study (as participatory as possible) DDR improve coordination innovative methodological tool for timely measurement and analysis of indicators networking and support of local NGOs and CSOs enhanced food security information conflict impact assessment disaster preparedness strategy/contingency planning effective advocacy for resource mobilization and FS policy/political action coordination well-differentiated roles for different actors detailed contextual analysis low profile interventions experienced staff within agencies funding and resources creativity and flexibility clear objectives and clear timeframes making use of existing information local consultation in needs assessment safety and security of agency staff peacekeeping forces with protection mandate willingness to work with non-traditional actors awareness of and learning from past interventions 3 capacity to analyse livelihoods context operational capacity to act on information in different timeframes flexibility of response timetable (donors) coordination approach between actors capacity to analyse contexts in conflict ethical checklist institutional memory continuity of agency presence through local staff attention to ethnic/clan support networks

When there is the threat of violent conflict

Acute crisis involving violent conflict

Lulls in conflict of uncertain duration

When longer lasting peace seems likely support to civil society with interest in sustainable peace reliable and shared information and analysis (baseline variation, shocks, including language) accepted link with protection official commitment to peace process and willingness for external input - e.g. dismantling the war economy mechanisms for dispute resolution over resources/assets constructive and critical participative dialogue at all levels agreed food security strategy stakeholder coordination

Whats needed to make it work?

accountability of all players institutional capacity building transparent management of grain reserves large scale outreach projects empowerment of the vulnerable targeting the most vulnerable better early warning indicators appropriate incentives both economic and institutional understanding the context/dynamics

Multidimensional crisis without conflict INGOs and Red Cross government UN economic actors winners and losers ghost actors (including 'spooks') other governments regional and others civil society - CSO, CBO, NGO, political IFIs national and international media donors local authorities international agencies vulnerable elements in society - participation in peacekeeping, definition of priorities warring parties civil society actors affected communities ethnic/religious/clan networks external - donors, UN, development agencies, longer term reconstruction perspectives media peacekeeping forces missions, churches and religious institutions warlords UN child soldiers external players - e.g. Kagame Museveni (Ituri) human rights agencies NGOs IDPs settled population govt & alternative authorities military, merchants, politicians multinationals donors

When there is the threat of violent conflict

Acute crisis involving violent conflict

Lulls in conflict of uncertain duration

When longer lasting peace seems likely armed actors reinsertion sustainability private sector sustainable market entry points throughout consider gender, ethnicity, religion, traditions national authorities - need rebuilding, an essential element of conflict resolution local associations, civil society needs, priorities, sustainability

What actors are involved?

min of planning/finance parliament/mps min agriculture national AIDS council private sector NGO/civil society research communities farmers associations, producers, other lobby groups donors private foundations technical associations and organizations (relief, development & humanitarian)